sample.jsonl

{"id": "bb-05244", "exam": "CompTIA Security+", "exam_code": "SY0-701", "product": "Study Guide", "question": "An IoT sensor network must authenticate devices with minimal overhead. Which is the BEST fit?", "options": {"A": "Full PKI with RSA-4096 for every sensor", "B": "Lightweight certs or pre-shared keys with AES", "C": "Disabling all authentication for speed", "D": "Requiring MFA push notifications per sensor"}, "answer": "B", "explanation": "Constrained devices need lightweight crypto. Full RSA-4096 PKI is too heavy; no auth is never the answer; human MFA per sensor is absurd. Match the control to the constraint — classic scenario logic.", "domain": null}
{"id": "bb-03380", "exam": "Family Nurse Practitioner", "exam_code": "FNP", "product": "Practice Pack", "question": "During a well-child visit, a 4-year-old has bruises in various stages of healing on the buttocks and back, and flinches when the father raises his voice. The father says the child \"bruises easily.\" The FNP's legal and ethical obligation is to:", "options": {"A": "Confront the father and accuse him of abuse", "B": "Document the findings but take no action without proof", "C": "Ask the father for permission before doing anything", "D": "Report the suspected child abuse to the state child protective agency as required by mandatory reporting law"}, "answer": "D", "explanation": "Suspicious bruising patterns (buttocks, back, multiple ages) with a fearful child meet the threshold for suspected child abuse, and every state designates health care providers as mandatory reporters — the obligation is to report reasonable suspicion to child protective services, not to prove abuse or obtain parental permission first. Confronting or accusing the alleged abuser endangers the child and compromises the investigation. Documenting without reporting violates the law and leaves the child unprotected. Know your state's reporting hotline and process in advance; immunity protects good-faith reporters. Ensure the child's immediate safety during the visit and arrange appropriate medical evaluation of the injuries.", "domain": null}
{"id": "bb-01147", "exam": "Microsoft Azure Administrator", "exam_code": "AZ-104", "product": "Study Guide", "question": "Two virtual machines exchange thousands of small messages per second and need the lowest possible network latency between them. What should you configure?", "options": {"A": "An availability set with 3 fault domains", "B": "A virtual machine scale set", "C": "Accelerated networking on a larger VM size", "D": "A proximity placement group for both VMs"}, "answer": "D", "explanation": "A proximity placement group co-locates VMs on the same (or nearby) physical hardware to minimize network latency — the standard answer for latency-sensitive pairs. Availability sets (A) spread VMs apart for resilience, which is the opposite goal; accelerated networking (C) helps throughput but not placement.", "domain": null}
{"id": "bb-03327", "exam": "Family Nurse Practitioner", "exam_code": "FNP", "product": "Practice Pack", "question": "A 45-year-old woman reports episodic right upper quadrant pain 30-60 minutes after fatty meals, lasting 1-2 hours and resolving completely. She has no fever, no Murphy sign, and normal vital signs. Between episodes she feels well. The most likely diagnosis is:", "options": {"A": "Acute cholecystitis", "B": "Biliary colic", "C": "Peptic ulcer disease", "D": "Acute hepatitis"}, "answer": "B", "explanation": "Episodic postprandial RUQ pain that resolves completely between attacks, without fever or a Murphy sign, is biliary colic — transient cystic duct obstruction by a gallstone that passes. Acute cholecystitis, by contrast, causes persistent pain beyond 6 hours with fever, leukocytosis, and a positive Murphy sign (inspiratory arrest on palpation). Peptic ulcer disease produces epigastric burning related to meals or NSAIDs, not RUQ colic. Acute hepatitis presents with malaise, jaundice, and markedly elevated transaminases rather than episodic pain. Confirm biliary colic with right upper quadrant ultrasound showing gallstones, and refer for elective cholecystectomy if episodes recur.", "domain": null}
{"id": "bb-00994", "exam": "AWS Certified AI Practitioner", "exam_code": "AIF-C01", "product": "Practice Pack", "question": "Which scenario is a classic use case for machine learning?", "options": {"A": "Forecasting future product demand based on historical sales patterns", "B": "Sorting a list of names alphabetically", "C": "Storing customer records in a database", "D": "Calculating exact payroll taxes using fixed formulas"}, "answer": "A", "explanation": "ML excels at learning patterns from historical data to predict future outcomes, such as demand forecasting. Fixed calculations, sorting, and storage are deterministic tasks that do not require learning from data.", "domain": "Domain 1"}
{"id": "bb-02055", "exam": "ISC2 CCSP", "exam_code": "CCSP", "product": "Practice Pack", "question": "Cloud Platform A company replicates its database continuously to a second region but keeps application servers as machine images that are launched only during failover. Which DR approach is this?", "options": {"A": "Pilot light", "B": "Hot standby", "C": "Cold site", "D": "Backup and restore"}, "answer": "A", "explanation": "Pilot light (A) keeps a minimal core — typically the replicated database — running in the recovery region, with remaining servers provisioned from images at failover time: faster than backup-and-restore, cheaper than hot standby. Hot standby (B) runs full capacity continuously. Cold site (C) has nothing provisioned. Backup and restore (D) is the slowest approach.", "domain": "Cloud Platform"}
{"id": "bb-02584", "exam": "Certified Kubernetes Application Developer", "exam_code": "CKAD", "product": "Study Guide", "question": "You create a Role granting ConfigMap reads and a RoleBinding for ServiceAccount reader in namespace team-a. A pod in namespace team-b running as system:serviceaccount:team-a:reader still gets Forbidden listing ConfigMaps in team-b. Why?", "options": {"A": "ServiceAccounts cannot be used outside their home namespace at all", "B": "RoleBindings grant permissions only within their own namespace — the binding in team-a says nothing about team-b; you need a second RoleBinding in team-b (or a ClusterRoleBinding)", "C": "The Role's verbs are wrong — list requires a ClusterRole", "D": "The token is namespace-bound and stops working in team-b"}, "answer": "B", "explanation": "RBAC grants are anchored to the binding's namespace: a RoleBinding in team-a authorizes actions only against resources in team-a, regardless of where the subject's pods run. To authorize the same SA in team-b, create a RoleBinding there too (referencing the team-a ServiceAccount as the subject) or use a ClusterRoleBinding for cluster-wide access. ServiceAccounts are perfectly usable cross-namespace (A is wrong); only the grant's scope was misunderstood.", "domain": null}
{"id": "bb-01964", "exam": "Adult CCRN", "exam_code": "CCRN", "product": "Study Guide", "question": "ABG: pH 7.28, PaCO2 55, HCO3 25. This represents:", "options": {"A": "Metabolic acidosis", "B": "Respiratory acidosis", "C": "Metabolic alkalosis", "D": "Respiratory alkalosis"}, "answer": "B", "explanation": "Low pH with elevated CO2 and near-normal bicarbonate is respiratory acidosis — hypoventilation. Improve ventilation; treat the cause.", "domain": null}
{"id": "bb-00036", "exam": "Google Professional ML Engineer", "exam_code": null, "product": "Practice Pack", "question": "A pipeline must automatically validate incoming training data: detect schema changes, anomalous value distributions, and missing features before training starts — failing fast if the data is bad. Which TFX components implement this?", "options": {"A": "Trainer and Pusher, which validate the model before deployment", "B": "Transform and Trainer, which clean the data during preprocessing", "C": "StatisticsGen, SchemaGen, and ExampleValidator, which compute stats, infer/verify schema, and flag anomalies", "D": "Evaluator and InfraValidator, which check the serving infrastructure"}, "answer": "C", "explanation": "C is correct because the TFX data-validation trio is purpose-built for this gate: StatisticsGen computes data statistics, SchemaGen infers or holds the expected schema, and ExampleValidator compares new data against the schema to flag skew, drift, and anomalies — failing the pipeline before wasted training. A is wrong: Trainer and Pusher train and deploy; they do not validate input data. B is wrong: Transform applies preprocessing but does not detect schema or distribution anomalies. D is wrong: Evaluator validates model quality and InfraValidator checks serving binaries — neither inspects raw training data.", "domain": null}
{"id": "bb-05439", "exam": "TEAS 7", "exam_code": "TEAS-7", "product": "Study Guide", "question": "Solve: 2(x − 3) = 14.", "options": {"A": "7", "B": "9", "C": "10", "D": "11"}, "answer": "C", "explanation": "2x − 6 = 14 → 2x = 20 → x = 10.", "domain": null}
{"id": "bb-00128", "exam": "Google Professional ML Engineer", "exam_code": null, "product": "Study Guide", "question": "A marketing team has no labeled data but wants to segment 2 million customers in BigQuery into distinct groups for targeted campaigns. Which BigQuery ML model type should they use?", "options": {"A": "ARIMA_PLUS", "B": "Logistic regression", "C": "Matrix factorization", "D": "K-means clustering"}, "answer": "D", "explanation": "K-means is BigQuery ML's unsupervised clustering model — it groups customers by similarity with no labels required, exactly what segmentation needs. Logistic regression (B) is supervised and needs a label. Matrix factorization (C) needs user-item interactions for recommendations. ARIMA_PLUS (A) forecasts time series. When the ask is 'group unlabeled rows,' k-means is the answer.", "domain": null}
{"id": "bb-00248", "exam": "HashiCorp Terraform Associate", "exam_code": "004", "product": "Practice Pack", "question": "A team moves their state from local files to an S3 backend by updating the `backend` block. What must they run next, and what does it do?", "options": {"A": "`terraform apply` — it recreates all infrastructure in the new backend", "B": "`terraform init -migrate-state` — it copies the existing state to the new backend so nothing is recreated", "C": "`terraform state push` — it uploads the configuration files", "D": "Nothing — Terraform detects the backend change automatically on the next plan ByteBar · Terraform Associate (004) Practice Pack — for personal use only 20"}, "answer": "B", "explanation": "Changing backends requires init -migrate-state, which moves the state file to the new home; resources are adopted, not recreated. A would be catastrophic and wrong. C uploads state but does not reconfigure the backend. D is false — Terraform prompts for migration via init. ByteBar · Terraform Associate (004) Practice Pack — for personal use only 38", "domain": "Domain 6"}
{"id": "bb-00282", "exam": "HashiCorp Terraform Associate", "exam_code": "004", "product": "Practice Pack", "question": "An organization has 40 workspaces that all need the same AWS credentials and region variables. What is the DRY way to manage this in HCP Terraform?", "options": {"A": "Copy the variables into each workspace by hand", "B": "Variable sets — define the shared variables once and attach the set to all 40 workspaces (or their project)", "C": "Store them in `terraform.tfvars` in each workspace's repo", "D": "Variable sets only work for a single workspace, so this is impossible ByteBar · Terraform Associate (004) Practice Pack — for personal use only 25"}, "answer": "B", "explanation": "project) Variable sets are shared variable collections attached across workspaces or projects — define once, inherit everywhere. A is the manual anti-pattern. C works for CLI runs but misses HCP's native mechanism. D is false; multi-workspace attachment is the point.", "domain": "Domain 8"}
{"id": "bb-03179", "exam": "CompTIA SecAI+", "exam_code": "CY0-001", "product": "Study Guide", "question": "By submitting many queries to a facial-recognition API and analyzing the confidence scores returned, an attacker reconstructs recognizable images of people from the training set. What is this attack?", "options": {"A": "Model inversion, because outputs are used to reconstruct sensitive inputs the model was trained on", "B": "Membership inference, because the attacker only learns whether a record was used", "C": "Model extraction, because the attacker builds a copy of the API", "D": "Evasion, because the attacker fools the model with perturbed inputs"}, "answer": "A", "explanation": "Model inversion is correct because the attacker works backward from the model's outputs to reconstruct the sensitive input data (faces) the model learned, exploiting the information the model memorized. Membership inference is wrong because that only answers whether a known record was in the training set; here the attacker recovers unknown faces. Model extraction is wrong because the goal is data recovery, not cloning model behavior. Evasion is wrong because the attacker is not trying to cause misclassification with perturbed inputs.", "domain": null}
{"id": "bb-03263", "exam": "Docker Certified Associate", "exam_code": "DCA", "product": "Practice Pack", "question": "An MKE upgrade is scheduled. The runbook's first step is a backup with --passphrase. The second step is verifying the backup. Why is backup verification critical for MKE specifically?", "options": {"A": "Verification is only needed for compliance audits", "B": "MKE backups are unencrypted and often corrupt", "C": "MKE backups can't be verified; the step is ceremonial", "D": "An MKE backup contains Swarm state, configs, and access control — and it's encrypted with the passphrase. An untested backup (wrong passphrase recorded, truncated file) is discovered only at restore time, when it's too late. Verify by test-restoring or at minimum validating the archive and passphrase record ByteBar · DCA Practice Pack — for personal use only 18 Domain 4: Networking — 15 questions"}, "answer": "D", "explanation": "passphrase. An untested backup (wrong passphrase recorded, truncated file) is discovered only at restore time, when it's too late. Verify by test-restoring or at minimum validating the archive and passphrase record The nightmare scenario: upgrade fails, you reach for the backup, and the recorded passphrase is wrong or the file is truncated — the backup is cryptographically useless and there's no recovery. MKE backups bundle the cluster's identity (Swarm Raft state, configs, grants), so their integrity IS your disaster-recovery plan. Best practice: store the passphrase in a vault (separate from the backup), verify archives after creation, and periodically test-restore into a lab. The exam loves the 'lose the passphrase = lose the backup' fact — this is its operational consequence.", "domain": "Domain 3"}
{"id": "bb-04121", "exam": "PMHNP-BC", "exam_code": "PMHNP-BC", "product": "Study Guide", "question": "A patient repeats the interviewer's questions back verbatim instead of answering. This is:", "options": {"A": "Echolalia", "B": "Neologism", "C": "Word salad", "D": "Clang association"}, "answer": "A", "explanation": "Echolalia is the automatic repetition of another's speech — seen in autism, catatonia, and some psychoses. Neologisms (B) are invented words; word salad (C) is incoherent mixing; clang associations (D) link words by sound, not meaning. Document exactly what you observe.", "domain": null}
{"id": "bb-04127", "exam": "PMHNP-BC", "exam_code": "PMHNP-BC", "product": "Study Guide", "question": "A patient stabilized on clozapine stops smoking. What should the PMHNP anticipate?", "options": {"A": "Immediate need to stop clozapine", "B": "Decreased clozapine levels requiring a dose increase", "C": "Increased clozapine levels with toxicity risk", "D": "No change in clozapine metabolism"}, "answer": "C", "explanation": "Tobacco smoke induces CYP1A2, which metabolizes clozapine. Quitting raises clozapine levels — monitor for toxicity (sedation, hypersalivation, seizures) and consider dose reduction.", "domain": null}
{"id": "bb-02268", "exam": "Certified Kubernetes Administrator", "exam_code": "CKA", "product": "Practice Pack", "question": "Pods must run only on nodes with SSDs, labeled disktype=ssd. The requirement may later grow to complex rules (ssd OR nvme, but never on spot nodes). Which mechanism should you use today to stay future-proof?", "options": {"A": "Pod anti-affinity against non-SSD nodes", "B": "nodeSelector: {disktype: ssd} — it supports OR and NOT operators", "C": "nodeAffinity with requiredDuringSchedulingIgnoredDuringExecution — it supports matchExpressions (In, NotIn, Exists) for complex rules, unlike nodeSelector", "D": "A taint on all non-SSD nodes"}, "answer": "C", "explanation": "NotIn, Exists) for complex rules, unlike nodeSelector nodeAffinity's matchExpressions support In/NotIn/Exists/Gt/Lt, covering the future OR/NOT requirements; nodeSelector only does exact-match AND. Taints invert the model awkwardly, and anti-affinity is about pod colocation, not node attributes.", "domain": "Domain 2"}
{"id": "bb-00776", "exam": "Google Associate Cloud Engineer", "exam_code": "ACE", "product": "Practice Pack", "question": "How long are logs retained in the default _Default log bucket unless the retention is changed?", "options": {"A": "7 days", "B": "7 years", "C": "10 years", "D": "30 days"}, "answer": "D", "explanation": "The _Default bucket retains logs for 30 days by default, configurable up to 3,650 days. Longer compliance retention typically uses sinks to Cloud Storage or BigQuery.", "domain": "Section 3"}
{"id": "bb-01551", "exam": "Microsoft Azure Fundamentals", "exam_code": "AZ-900", "product": "Study Guide", "question": "After a regional outage, an insurer restores its claims system from backups in a secondary region within hours. Which benefit does this demonstrate?", "options": {"A": "Reliability", "B": "Scalability", "C": "Predictability", "D": "Agility"}, "answer": "A", "explanation": "Reliability means systems recover from failures and keep operating — disaster recovery and backups are the textbook examples. Scalability handles more load; agility is speed of change. BYTEBAR · SWEET IT SKILLS, NO FLUFF ByteBar • AZ-900 Study Guide", "domain": "Domain 1"}
{"id": "bb-03870", "exam": "NCLEX-RN", "exam_code": "NCLEX-RN", "product": "Practice Pack", "question": "The nurse teaches diet to a client with chronic kidney disease. Which statement indicates correct understanding?", "options": {"A": "I can eat unlimited protein to stay strong.", "B": "Bananas and oranges are the best snacks.", "C": "I will limit protein, potassium, phosphorus, sodium, and fluids as prescribed.", "D": "I should drink at least 4 liters of water daily."}, "answer": "C", "explanation": "Failing kidneys cannot clear protein waste, potassium, phosphorus, or excess fluid, so the renal diet restricts protein to reduce uremic waste, limits potassium and phosphorus to prevent dysrhythmias and bone disease, controls sodium and fluids to manage blood pressure and edema — all individualized to lab values. Option A is wrong because excess protein accelerates waste accumulation. Option B is wrong because bananas and oranges are high in potassium and dangerous in CKD. Option D is wrong because most CKD clients require fluid restriction, not large-volume intake.", "domain": null}
{"id": "bb-00720", "exam": "Google Associate Cloud Engineer", "exam_code": "ACE", "product": "Practice Pack", "question": "A company wants to analyze detailed, line-item billing data with SQL queries. What should they set up?", "options": {"A": "A billing export to BigQuery", "B": "A billing budget", "C": "A Cloud Logging sink", "D": "A billing export to Cloud Storage only"}, "answer": "A", "explanation": "The BigQuery billing export loads detailed usage and cost data into tables you can query with SQL. The Cloud Storage export produces flat files, and budgets only trigger alerts.", "domain": "Section 1"}
{"id": "bb-00799", "exam": "Google Associate Cloud Engineer", "exam_code": "ACE", "product": "Practice Pack", "question": "You try to create a Cloud SQL instance with gcloud but get an error saying the API is not enabled. What is the fix?", "options": {"A": "Request a quota increase", "B": "Enable the Cloud SQL Admin API in the project", "C": "Create a new billing account", "D": "Grant yourself the Owner role"}, "answer": "B", "explanation": "Google Cloud APIs must be enabled per project before their services can be used. Quotas, billing, and IAM roles do not substitute for enabling the API.", "domain": "Section 1"}
{"id": "bb-01643", "exam": "ISC2 Certified in Cybersecurity", "exam_code": "CC", "product": "Practice Pack", "question": "All 200 customer-service reps receive identical access to the ticketing system based on their job function. This is:", "options": {"A": "DAC", "B": "MAC", "C": "RBAC", "D": "ABAC"}, "answer": "C", "explanation": "Role-Based Access Control (RBAC) — access follows the job role. Everyone in the role gets the same permissions, which makes joiners/movers/leavers easy to manage consistently.", "domain": null}
{"id": "bb-03800", "exam": "NCLEX-RN", "exam_code": "NCLEX-RN", "product": "Practice Pack", "question": "A client taking digoxin has a serum potassium level of 2.9 mEq/L. The nurse understands the client is at increased risk for:", "options": {"A": "Metabolic alkalosis.", "B": "Digoxin toxicity.", "C": "Digoxin underdosing.", "D": "Hypercalcemia."}, "answer": "B", "explanation": "Hypokalemia increases the binding of digoxin to myocardial cells, so even a therapeutic digoxin level can become toxic when potassium is low — the nurse watches for anorexia, nausea, vision changes, and dysrhythmias. Option C is wrong because the problem is exaggerated drug effect, not too little drug. Option D is wrong because calcium is not directly altered by this interaction. Option A is wrong because metabolic alkalosis is not the consequence; the danger is digoxin toxicity, which is why potassium is replaced and the provider is notified before the next dose.", "domain": null}
{"id": "bb-04695", "exam": "Microsoft SC-500", "exam_code": "SC-500", "product": "Practice Pack", "question": "Contoso designates a Microsoft Entra group as the SQL logical server's Entra admin. Six months later, every member of that group leaves the company and the group is deleted. What happens to Entra-based database access?", "options": {"A": "Existing contained users keep working; only new user provisioning breaks until a new admin is designated.", "B": "All Entra authentication to the server stops immediately because contained users are deleted with the group.", "C": "The server automatically promotes the subscription owner to Entra admin.", "D": "Nothing breaks — the admin designation persists on the deleted group's object ID and continues to authorize new users."}, "answer": "A", "explanation": "Existing contained users keep working; only new user provisioning breaks until a new admin is designated. — The Entra admin is needed to provision Entra principals (CREATE USER ... FROM EXTERNAL PROVIDER); already-provisioned contained users authenticate directly against Entra ID and are unaffected. With the admin group gone, no new Entra users can be created or dropped until a new admin is designated. Nothing is auto-promoted (C is false), contained users are not deleted with the group (B is false), and a deleted group's object ID authorizes nobody (D is false).", "domain": null}
{"id": "bb-05178", "exam": "CompTIA Security+", "exam_code": "SY0-701", "product": "Practice Pack", "question": "Two parties need to agree on an encryption key over an insecure channel with no prior shared secret. Use:", "options": {"A": "AES with a pre-shared key", "B": "Diffie-Hellman", "C": "MD5", "D": "HMAC"}, "answer": "B", "explanation": "Diffie-Hellman = secure key agreement between strangers. Then use the agreed key with AES for bulk encryption.", "domain": null}
{"id": "bb-04622", "exam": "AWS Solutions Architect Associate", "exam_code": "SAA-C03", "product": "Practice Pack", "question": "A data lake bucket must never be publicly accessible, and the company wants a guardrail against future misconfiguration. Which control is MOST effective?", "options": {"A": "Enable S3 Block Public Access at the account and bucket level", "B": "Name the bucket with a random string", "C": "Rely on the default ACLs alone", "D": "Enable S3 server access logging"}, "answer": "A", "explanation": "Block Public Access overrides any policy or ACL that would grant public access, at both account and bucket scope. Obscure names, default ACLs, and access logging don't prevent a future policy mistake from exposing the bucket.", "domain": "Domain 3"}
{"id": "bb-02302", "exam": "Certified Kubernetes Administrator", "exam_code": "CKA", "product": "Study Guide", "question": "A CronJob with its 2 AM run during a startingDeadlineSeconds: 100 missed control-plane outage. At 6 AM no Job exists and none will be created for that slot. Why?", "options": {"A": "CronJob schedules are evaluated only at the exact minute — the controller never does catch-up", "B": "concurrencyPolicy: Forbid blocks all future runs after one miss", "C": "The schedule field uses the worker node's local timezone", "D": "A missed start older than startingDeadlineSeconds is counted as failed and never started; the CronJob resumes at the next scheduled tick — check LastScheduleTime and controller logs, and raise the deadline if catch-up matters"}, "answer": "D", "explanation": "bounds how late a missed start can still fire; beyond it the startingDeadlineSeconds start is skipped, not retried, and the CronJob resumes at the next cron tick. The controller does attempt catch-up within the deadline.", "domain": null}
{"id": "bb-03125", "exam": "CompTIA SecAI+", "exam_code": "CY0-001", "product": "Study Guide", "question": "A SOC wants an email filter that labels incoming messages as either phishing or legitimate. The team has 10,000 historical emails, each already tagged by analysts as phishing or legitimate. Which machine learning approach fits this task?", "options": {"A": "Unsupervised learning, because the model should discover email groupings without any guidance", "B": "Supervised learning, because the model learns the mapping from labeled examples to the correct category", "C": "Reinforcement learning, because the model should earn rewards for each email it classifies", "D": "Federated learning, because the emails must be encrypted before any training can occur"}, "answer": "B", "explanation": "Supervised learning is correct because the training data is labeled: every email already carries the right answer (phishing or legitimate), and the model learns the mapping from email features to those labels so it can classify new mail. Unsupervised learning is wrong because it works on unlabeled data to find hidden structure, not to predict known categories. Reinforcement learning is wrong because it learns through trial-and-error rewards in an environment, not from labeled examples. Federated learning is wrong because it describes training across distributed devices without centralizing data, which has nothing to do with whether the data is labeled.", "domain": null}
{"id": "bb-00972", "exam": "AWS Certified AI Practitioner", "exam_code": "AIF-C01", "product": "Practice Pack", "question": "What is Amazon Bedrock?", "options": {"A": "A video streaming service", "B": "A fully managed service for building generative AI applications with foundation models through a single API, without managing infrastructure", "C": "A physical data center location", "D": "A database migration tool"}, "answer": "B", "explanation": "Amazon Bedrock is a fully managed service that provides access to high-performing foundation models from Amazon and third-party providers through one API, with serverless inference - no infrastructure to manage.", "domain": "Domain 3"}
{"id": "bb-02280", "exam": "Certified Kubernetes Administrator", "exam_code": "CKA", "product": "Practice Pack", "question": "You create a Service of type ClusterIP for a database that only in-cluster apps should reach. A teammate suggests NodePort so they can test from their laptop. Why is ClusterIP the right call here?", "options": {"A": "ClusterIP automatically encrypts traffic; NodePort does not", "B": "NodePort does not work for databases", "C": "ClusterIP is faster because it bypasses kube-proxy", "D": "ClusterIP exposes the service only on an internal virtual IP reachable within the cluster; NodePort would open a port on every node's public interface, widening exposure unnecessarily"}, "answer": "D", "explanation": "open a port on every node's public interface, widening exposure unnecessarily ClusterIP allocates a virtual IP routable only inside the cluster — the correct scope for an internal database. NodePort opens a high port on ALL nodes' interfaces, which is both a security widening and unnecessary for in-cluster consumers. Neither type encrypts traffic, and both work for any TCP/UDP workload.", "domain": "Domain 3"}
{"id": "bb-00281", "exam": "HashiCorp Terraform Associate", "exam_code": "004", "product": "Practice Pack", "question": "An external tagging tool manages the `tags` on instances, and every `terraform plan` wants to revert its changes. How do you make Terraform leave the tags alone?", "options": {"A": "`lifecycle { ignore_changes = [tags] }` — Terraform stops tracking drift on those attributes", "B": "Remove the `tags` argument from the configuration entirely", "C": "`lifecycle { prevent_destroy = true }`", "D": "Run every plan with `-refresh=false`"}, "answer": "A", "explanation": "ignore_changes tells Terraform to accept external modifications to listed attributes instead of reverting them. B does not help — Terraform would still manage tags as empty. C blocks destruction, unrelated to attribute drift. D disables refresh globally, hiding all drift — dangerous and wrong.", "domain": "Domain 4"}
{"id": "bb-04951", "exam": "Subnetting Mastery", "exam_code": "SUBNET", "product": "Practice Pack", "question": "What is the decimal number 224 in binary?", "options": {"A": "10100000", "B": "11100000", "C": "11010000", "D": "11110000"}, "answer": "B", "explanation": "224 = 128 + 64 + 32 → 11100000.", "domain": null}
{"id": "bb-00570", "exam": "CompTIA A+", "exam_code": "220-1201/220-1202", "product": "Practice Pack", "question": "What is the purpose of a firewall's implicit deny?", "options": {"A": "Allow all traffic by default", "B": "Block everything not explicitly permitted", "C": "Log all connections", "D": "Encrypt traffic"}, "answer": "B", "explanation": "Implicit deny = default-drop. Rules allow specific traffic; everything else is blocked — the secure default.", "domain": null}
{"id": "bb-03763", "exam": "CompTIA Network+", "exam_code": "N10-009", "product": "Study Guide", "question": "A switch port is configured to learn the first connected device’s MAC address, keep it, and shut the port down if a different device connects. Which feature is this? ByteBar · Network+ N10-009 Study Guide — for personal use only 22 BYTEBAR · SWEET IT SKILLS, NO FLUFF", "options": {"A": "DHCP snooping", "B": "Dynamic ARP Inspection", "C": "Port security with sticky MAC and shutdown violation mode", "D": "802.1Q trunking"}, "answer": "C", "explanation": "Port security limits MACs per port; “sticky” learns and locks the first MAC seen; shutdown mode disables the port on violation (vs. protect = drop silently, restrict = drop + log). The classic unauthorized-device defense. Exam-day playbook • PBQs first. They’re heavy and slow — subnetting, CLI config, drag-and-drop topologies. Get them done while your brain is fresh. • 90 minutes / 90 questions = 1 minute each. If a question takes longer, flag it and move on — the exam lets you review flagged questions at the end. • Read the LAST sentence first. It tells you what’s actually being asked; then skim the scenario for constraints. • Eliminate two, choose between two. CompTIA loves two plausible answers — pick the one that BEST fits all stated constraints (cheapest? most secure? first step?). • Watch for scope words: ‘MOST secure’, ‘FIRST step’, ‘BEST’, ‘LEAST expensive’ — they change the answer. • Think in layers. Start at Layer 1 (is it plugged in? link light?) and work up — half of all troubleshooting questions are solved at the physical layer. • Never leave blanks. There’s no penalty for wrong answers. • Sleep and logistics: two forms of ID, arrive 30 min early, and don’t cram the morning of — trust the prep. You’ve got the knowledge. Now go earn the cert. ByteBar · Sweet IT Skills, No Fluff Independent study resource · Not affiliated with CompTIA, Inc. ByteBar · Network+ N10-009 Study Guide — for personal use only 23 BYTEBAR · SWEET IT SKILLS, NO FLUFF Practice Questions 31-100 70 new scenario-based questions covering all five N10-009 domains. Answer closed-book, then read every explanation - including the ones you got right.", "domain": null}
{"id": "bb-05378", "exam": "TEAS 7", "exam_code": "TEAS-7", "product": "Practice Pack", "question": "Which structures do ligaments connect?", "options": {"A": "Muscle to bone", "B": "Bone to bone", "C": "Muscle to muscle", "D": "Bone to skin"}, "answer": "B", "explanation": "B is correct because ligaments are dense bands of connective tissue that join one bone to another across a joint, stabilizing it — for example, the ACL connects the femur to the tibia. A describes tendons, the classic mix-up, since tendons attach muscle to bone. C is wrong because muscles do not connect to each other via ligaments. D is wrong because no ligament attaches bone to skin; skin is anchored by fascia and subcutaneous tissue.", "domain": null}
{"id": "bb-02782", "exam": "AWS Certified Cloud Practitioner", "exam_code": "CLF-C02", "product": "Study Guide", "question": "The finance team needs the most granular billing data available — hourly, line-item usage per resource — to feed a custom chargeback system. Which provides it? BYTEBAR · SWEET IT SKILLS, NO FLUFF", "options": {"A": "The Cost and Usage Report (CUR)", "B": "AWS Cost Explorer", "C": "AWS Budgets", "D": "The AWS Pricing Calculator"}, "answer": "A", "explanation": "The CUR is the raw, most-detailed billing dataset, delivered to S3 for custom analysis. Cost Explorer visualizes and analyzes spend; Budgets alerts on thresholds; the Pricing Calculator estimates costs before you build — none give raw line-item data.", "domain": null}
{"id": "bb-02128", "exam": "ISC2 CCSP", "exam_code": "CCSP", "product": "Study Guide", "question": "A discovery tool scans databases, file shares, and email to find unprotected credit card numbers. Which statement about this data is TRUE?", "options": {"A": "Credit card numbers in databases are unstructured data", "B": "Metadata is irrelevant to data discovery", "C": "Only structured data can be classified", "D": "The tool must handle both structured (databases) and unstructured (email, documents) data"}, "answer": "D", "explanation": "Discovery must cover structured sources like databases and unstructured ones like email and files (D). (A) is wrong because database fields are structured data. (B) is wrong because metadata and tags drive classification and discovery. (C) is wrong because unstructured data is routinely classified.", "domain": null}
{"id": "bb-04725", "exam": "Microsoft SC-500", "exam_code": "SC-500", "product": "Practice Pack", "question": "Contoso's production networking resource group must be protected so that no one, including subscription Owners, can delete it, while Owners keep full management rights over everything else. What is the BEST solution?", "options": {"A": "Create an Azure Policy Deny assignment blocking the delete action on the resource group", "B": "Apply a CanNotDelete management lock on the production networking resource group", "C": "Remove the Owner role from all administrators on the subscription", "D": "Move the production networking resources to a separate subscription"}, "answer": "B", "explanation": "Apply a CanNotDelete management lock on the production networking resource group — A CanNotDelete management lock prevents deletion of the resource group and its resources by any principal, including Owners, while leaving all other management operations intact. A Deny policy effect can be circumvented by policy exemptions and is not the designed mechanism for deletion protection. C strips needed administrative capability. D reorganizes billing and access boundaries without actually preventing deletion.", "domain": null}
{"id": "bb-00926", "exam": "AWS Certified AI Practitioner", "exam_code": "AIF-C01", "product": "Practice Pack", "question": "Amazon Polly is used to:", "options": {"A": "Forecast sales", "B": "Transcribe meetings", "C": "Convert text into lifelike speech", "D": "Detect objects in images"}, "answer": "C", "explanation": "Amazon Polly is a text-to-speech service that turns text into natural-sounding spoken audio, with many voices and languages.", "domain": "Domain 3"}
{"id": "bb-04578", "exam": "Red Hat RHCSA/RHCE", "exam_code": "RHCSA", "product": "Study Guide", "question": "You want every playbook run from ~/projects/web/ to use the inventory ~/projects/web/inventory, connect as user devops, and escalate privileges by default without passing CLI flags. Which ansible.cfg entries accomplish this?", "options": {"A": "[defaults] hosts = ./inventory ssh_user = devops sudo = true", "B": "[connection] inventory = ./inventory remote_user = devops [escalation] become = 1", "C": "[defaults] inventory = ./inventory remote_user = devops [privilege_escalation] become = True", "D": "[inventory] file = ./inventory user = devops become = yes"}, "answer": "C", "explanation": "C is correct: under [defaults], 'inventory' points at the inventory file and 'remote_user' sets the SSH login user, while the [privilege_escalation] section's 'become = True' enables privilege escalation by default. D is wrong: the section/key names [inventory], 'file', and 'user' do not exist. A is wrong: 'hosts' and 'ssh_user' are not valid keys, and 'sudo = true' is obsolete legacy syntax. B is wrong: there is no [connection] inventory key and no [escalation] section; privilege escalation settings live under [privilege_escalation].", "domain": null}
{"id": "bb-02208", "exam": "Certified Kubernetes Administrator", "exam_code": "CKA", "product": "Practice Pack", "question": "A developer needs permission to list and delete pods, but ONLY in the dev namespace. Which RBAC approach follows least privilege?", "options": {"A": "Editing the kube-apiserver manifest to add the developer as a superuser", "B": "A ClusterRole with pods delete/list verbs bound with a RoleBinding in the dev namespace", "C": "A ClusterRole with * verbs bound with a ClusterRoleBinding to the developer", "D": "A Role with pods get/list/delete verbs in the dev namespace, bound with a RoleBinding to the developer"}, "answer": "D", "explanation": "A namespaced Role granting only the needed verbs, bound via a RoleBinding in that namespace, is least privilege. A ClusterRole bound with a RoleBinding in one namespace technically scopes to that namespace, but a ClusterRole is reusable cluster-wide and over-broad here; a ClusterRoleBinding would grant access everywhere. The API server has no superuser list to edit.", "domain": "Domain 1"}
{"id": "bb-02258", "exam": "Certified Kubernetes Administrator", "exam_code": "CKA", "product": "Practice Pack", "question": "For a highly available cluster, clients must reach the API servers through a single stable endpoint even if one control-plane node dies. Which component provides this, and where is it configured at install time?", "options": {"A": "A load balancer (or virtual IP such as kube-vip) in front of the API servers, set as --control-plane-endpoint during kubeadm init", "B": "CoreDNS — configured with an A record for the API", "C": "kube-proxy in IPVS mode, configured on each worker", "D": "An Ingress resource pointing at the API server Service"}, "answer": "A", "explanation": "--control-plane-endpoint during kubeadm init --control-plane-endpoint is the stable DNS name or IP (backed by an external LB or a virtual IP like kube-vip) that kubeadm bakes into kubeconfigs and certs; if a control-plane node dies, traffic fails over. CoreDNS serves in-cluster service DNS, not external API access; Ingress routes HTTP workloads, not the API; kube-proxy handles Service IPs on nodes.", "domain": "Domain 1"}
{"id": "bb-05109", "exam": "CompTIA Security+", "exam_code": "SY0-701", "product": "Practice Pack", "question": "A login page is cloned pixel-perfectly, but the URL is 'micros0ft-login.com'. This is:", "options": {"A": "Typosquatting / brand impersonation", "B": "A watering hole", "C": "Bluejacking", "D": "A replay attack"}, "answer": "A", "explanation": "Lookalike domain mimicking a trusted brand = typosquatting, a form of brand impersonation used in phishing.", "domain": null}
{"id": "bb-02542", "exam": "Certified Kubernetes Application Developer", "exam_code": "CKAD", "product": "Study Guide", "question": "helm install my-web ./chart fails with cannot re-use a name that is still in use, and helm list -A shows my-web as failed. What is the clean fix?", "options": {"A": "Run helm install with --force to overwrite the failed release", "B": "Delete the namespace the release was installed into", "C": "Wait — failed releases are garbage-collected automatically within an hour", "D": "helm uninstall my-web to remove the failed release, then install again (or use helm upgrade --install) ByteBar · Certified Kubernetes Application Developer (CKAD) Study Guide — for personal use only 66"}, "answer": "D", "explanation": "A failed release still holds its name in Helm's storage, blocking reuse. helm uninstall clears it so a fresh install can proceed; helm upgrade --install handles both cases idempotently. Failed releases are not auto-collected (C), --force (A) is not an install flag for this, and deleting the namespace (B) is drastic and leaves Helm storage inconsistent.", "domain": null}
{"id": "bb-00562", "exam": "CompTIA A+", "exam_code": "220-1201/220-1202", "product": "Practice Pack", "question": "What does BitLocker provide?", "options": {"A": "Network encryption", "B": "Full-disk encryption for Windows", "C": "Email filtering", "D": "Password management"}, "answer": "B", "explanation": "BitLocker encrypts entire Windows volumes — data-at-rest protection if the device is lost or stolen.", "domain": null}
{"id": "bb-02876", "exam": "CompTIA CySA+", "exam_code": "CS0-004", "product": "Practice Pack", "question": "You find a scheduled task on a server that runs a Base64-encoded script at logon, created by a user account that was disabled last month. What should you conclude?", "options": {"A": "Base64 encoding proves the script is safe", "B": "This is normal — disabled accounts routinely create scheduled tasks", "C": "This is a persistence mechanism and should be treated as a suspected compromise until proven otherwise", "D": "Scheduled tasks cannot be malicious"}, "answer": "C", "explanation": "otherwise Persistence via scheduled tasks with obfuscated payloads, tied to a disabled account, is a textbook attacker foothold (ATT&CK; T1053). Disabled accounts do not create tasks, and encoding is obfuscation, not a safety guarantee.", "domain": "Domain 1"}
{"id": "bb-04796", "exam": "Microsoft SC-500", "exam_code": "SC-500", "product": "Study Guide", "question": "Contoso connects its datacenter to Azure with ExpressRoute private peering. The architects state that traffic must never traverse the public internet and must be encrypted. Which statement is correct?", "options": {"A": "ExpressRoute private peering encrypts all traffic by default.", "B": "ExpressRoute traffic does not traverse the public internet, but it is not encrypted by default — add IPsec over ExpressRoute or MACsec on the physical links where encryption is required.", "C": "Private peering requires a VPN gateway to function.", "D": "NSGs attached to the ExpressRoute circuit filter the traffic."}, "answer": "B", "explanation": "ExpressRoute is a private Layer 3 connection that never touches the public internet, but Microsoft does not encrypt it by default; customers add IPsec tunnels over the peering or MACsec at the physical layer for confidentiality. Option A is the common misconception this question tests. A VPN gateway (C) is optional, not required. NSGs (D) apply to subnets and NICs, not to ExpressRoute circuits. Note that Microsoft peering (for Microsoft 365/Dynamics) is separate from private peering — keep the two peerings distinct in your design.", "domain": null}
{"id": "bb-04186", "exam": "CompTIA PenTest+", "exam_code": "PT0-003", "product": "Practice Pack", "question": "A critical vulnerability is publicly disclosed on Monday and the patch is applied on Friday. What is the Monday-to-Friday period called?", "options": {"A": "Dwell time", "B": "Scan interval", "C": "Service-level agreement", "D": "Window of exposure"}, "answer": "D", "explanation": "The window of exposure is the period between a vulnerability's disclosure and its remediation — the exact span during which attackers can exploit it, which is why rapid patching shrinks attacker opportunity. Dwell time measures how long an attacker remained undetected inside an environment, a different metric. Scan interval is simply how often assessments run. A service-level agreement is a contractual uptime commitment, unrelated to vulnerability timelines.", "domain": null}
{"id": "bb-02915", "exam": "CompTIA Cloud+", "exam_code": "CV0-004", "product": "Practice Pack", "question": "A company worries that if a physical disk is stolen from a datacenter, its data could be read. Which control BEST addresses this risk?", "options": {"A": "A longer password policy", "B": "Encryption at rest", "C": "A web application firewall", "D": "Disabling unused ports"}, "answer": "B", "explanation": "Encryption at rest protects stored data so a stolen disk yields only unreadable ciphertext without the keys. Password policies govern authentication, a WAF protects web traffic, and disabling ports reduces network attack surface — none protect data on a stolen disk.", "domain": "Domain 4"}
{"id": "bb-00319", "exam": "Cisco CCNA", "exam_code": "200-301", "product": "Practice Pack", "question": "Ansible automates network devices primarily over which protocol?", "options": {"A": "SNMP", "B": "SSH", "C": "Telnet", "D": "FTP"}, "answer": "B", "explanation": "Ansible is agentless over SSH, executing tasks defined in YAML playbooks.", "domain": "D5"}
{"id": "bb-03642", "exam": "CompTIA Network+", "exam_code": "N10-009", "product": "Practice Pack", "question": "An admin wants to inspect traffic to a server without interrupting it. Configuring the switch to copy traffic to a monitoring port is called:", "options": {"A": "Port mirroring (SPAN)", "B": "Port security", "C": "Trunking", "D": "NAT"}, "answer": "A", "explanation": "SPAN/mirroring copies frames to an analyzer port where Wireshark can capture them. Port security LIMITS MACs per port - the opposite job. Mirror to monitor; secure to restrict.", "domain": null}
{"id": "bb-02653", "exam": "AWS Certified Cloud Practitioner", "exam_code": "CLF-C02", "product": "Practice Pack", "question": "Which AWS CAF perspective covers organizational structure, roles, and the skills teams need for cloud adoption?", "options": {"A": "Governance", "B": "People", "C": "Security", "D": "Platform"}, "answer": "B", "explanation": "The People perspective addresses culture, org structure, leadership, and workforce skills. Governance covers policies and controls; Platform covers the landing-zone architecture. ByteBar · AWS Certified Cloud Practitioner (CLF-C02) Practice Pack — 100 Questions", "domain": "D1"}
{"id": "bb-02700", "exam": "AWS Certified Cloud Practitioner", "exam_code": "CLF-C02", "product": "Study Guide", "question": "A company must migrate 500 TB of archives to AWS, but its internet link is so slow that a network transfer would take months. Which service should it use?", "options": {"A": "AWS DataSync over the internet", "B": "AWS Snowball", "C": "S3 Transfer Acceleration", "D": "AWS Direct Connect"}, "answer": "B", "explanation": "Snowball = rugged physical devices shipped to you; load 500 TB locally and ship them back — built for petabyte-scale offline transfers. DataSync and Transfer Acceleration still need the network; Direct Connect gives a private link but doesn’t fix a slow one.", "domain": null}
{"id": "bb-04360", "exam": "CompTIA PenTest+", "exam_code": "PT0-003", "product": "Study Guide", "question": "You compromise a container running with --privileged and the host filesystem mounted. What is the primary risk?", "options": {"A": "The container runs slower", "B": "The container image is signed", "C": "The container restarts automatically", "D": "The container escapes to the host"}, "answer": "D", "explanation": "A privileged container with host mounts (especially / mounted or docker.sock) can break out to the host — the critical container escape finding. Auto-restart, image signing, and performance are unrelated to the escape risk.", "domain": null}
{"id": "bb-04208", "exam": "CompTIA PenTest+", "exam_code": "PT0-003", "product": "Practice Pack", "question": "An attacker enters a secure facility by walking closely behind a badged employee through the turnstile. What is this called?", "options": {"A": "Phishing", "B": "Piggybacking", "C": "Watering hole", "D": "Vishing"}, "answer": "B", "explanation": "Piggybacking/tailgating is physically following someone through a secured entry. Phishing is email-based, watering hole compromises frequented websites, vishing is voice phishing. Section 2 — Reconnaissance & Enumeration", "domain": null}
{"id": "bb-01194", "exam": "Microsoft Azure Administrator", "exam_code": "AZ-104", "product": "Study Guide", "question": "Your company has 12 subscriptions. You must ensure that no one can create virtual machines with more than 4 vCPUs in any subscription, and you want a single place to manage this rule. What should you do?", "options": {"A": "Create an Azure Policy assignment at the management group containing the 12 subscriptions, using the Deny effect", "B": "Assign the Contributor role at the management group with a custom NotAction", "C": "Apply a CanNotDelete lock on each subscription", "D": "Create a budget alert for each subscription"}, "answer": "A", "explanation": "Azure Policy governs what resource configurations are allowed, and assigning it at the management group scope covers all 12 subscriptions from one place. The Deny effect blocks non-compliant VM creation. RBAC (B) controls who can act, not what configurations are allowed. Locks (C) prevent accidental deletion or modification, not SKU selection. Budgets (D) only alert on spending.", "domain": null}
{"id": "bb-03937", "exam": "NCLEX-RN", "exam_code": "NCLEX-RN", "product": "Study Guide", "question": "Before a CT scan with IV contrast, the nurse’s priority assessment is to ask about:", "options": {"A": "Shoe size", "B": "Favorite foods", "C": "Exercise habits", "D": "Any prior reaction to IV contrast dye"}, "answer": "D", "explanation": "A prior contrast reaction is the real predictor of another — screen for it and premedicate per protocol. The shellfish/iodine ‘allergy’ link is a myth: shellfish allergy doesn’t predict contrast reactions, so a shrimp allergy alone shouldn’t cancel the scan. Also check kidney function and metformin use, and hydrate afterward to protect the kidneys. Favorite foods (B), exercise (C), and shoe size (A) are chart filler, not safety screening.", "domain": null}
{"id": "bb-00761", "exam": "Google Associate Cloud Engineer", "exam_code": "ACE", "product": "Practice Pack", "question": "Administrators need SSH and RDP access to VMs that have no public IP addresses, without deploying bastion hosts. What provides this securely?", "options": {"A": "Temporarily assigning public IPs", "B": "Cloud NAT", "C": "A VPN tunnel to each VM", "D": "Identity-Aware Proxy (IAP) TCP forwarding"}, "answer": "D", "explanation": "IAP TCP forwarding tunnels administrative protocols through Google's infrastructure with IAM-based access control — no bastions or public IPs needed. Cloud NAT only provides outbound connectivity.", "domain": "Section 4"}
{"id": "bb-03522", "exam": "ITIL 4 Foundation", "exam_code": "ITIL-4", "product": "Practice Pack", "question": "Key concepts A service offers exactly the functionality a customer needs, but it is frequently unavailable during business hours. In ITIL terms, the service has:", "options": {"A": "Utility but insufficient warranty", "B": "Warranty but no utility", "C": "Neither utility nor warranty", "D": "Value but no outcome"}, "answer": "A", "explanation": "Utility is fit for purpose — the functionality is right. Warranty is fit for use — available, reliable, and secure enough. Frequent outages are a warranty failure, not a utility failure.", "domain": "Key concepts"}
{"id": "bb-03893", "exam": "NCLEX-RN", "exam_code": "NCLEX-RN", "product": "Study Guide", "question": "A client is diagnosed with MRSA wound infection. The nurse implements:", "options": {"A": "Droplet precautions with a surgical mask", "B": "Airborne precautions with an N95 respirator", "C": "Standard precautions only", "D": "Contact precautions: gown, gloves, private room or cohort, dedicated equipment"}, "answer": "D", "explanation": "MRSA spreads by contact — gown and gloves, private or cohort room, dedicated equipment, hand hygiene. It’s not airborne (B) or droplet (A); standard precautions alone (C) miss the transmission route. Same playbook as VRE.", "domain": null}
{"id": "bb-03481", "exam": "Family Nurse Practitioner", "exam_code": "FNP", "product": "Study Guide", "question": "An 82-year-old on glyburide has had two episodes of sweating, confusion, and shakiness that resolved with juice. A1C is 7.1%. The best next step is:", "options": {"A": "Switch to a glucose-lowering agent with lower hypoglycemia risk", "B": "Add basal insulin", "C": "Continue glyburide — the A1C is at goal", "D": "Stop all diabetes medications"}, "answer": "A", "explanation": "Recurrent hypoglycemia on glyburide — a long-acting sulfonylurea — in an 82-year-old is a fall, fracture, and cardiac event waiting to happen; switch to an agent with minimal hypo risk (like a DPP-4 inhibitor). An A1C 'at goal' means nothing if getting there causes hypoglycemia. Adding insulin would worsen the hypo risk. Stopping everything abandons glycemic control instead of fixing the drug choice.", "domain": null}
{"id": "bb-02176", "exam": "ISC2 CCSP", "exam_code": "CCSP", "product": "Study Guide", "question": "Every person who handles a forensic disk image signs a log recording date, time, and purpose. Which principle does this uphold?", "options": {"A": "Data minimization", "B": "Order of volatility", "C": "Chain of custody", "D": "Legal hold"}, "answer": "C", "explanation": "Documented handling history preserves the chain of custody (C), which is required for evidence admissibility. Minimization (A) is wrong because it is a privacy principle. Order of volatility (B) is wrong because it governs collection sequencing. Legal hold (D) is wrong because it is the preservation duty, not the handling documentation.", "domain": null}
{"id": "bb-01376", "exam": "Microsoft AZ-802", "exam_code": "AZ-802", "product": "Study Guide", "question": "You need to delegate the ability to reset passwords for users in the Helpdesk OU to the help-desk team, without making them Domain Admins or Account Operators. What is the best approach?", "options": {"A": "Add the help-desk group to the Account Operators built-in group", "B": "Use the Delegation of Control Wizard on the Helpdesk OU to grant 'Reset user passwords and force password change at next logon'", "C": "Grant the help-desk group Full Control on the domain object", "D": "Make the help-desk team members local administrators on each domain controller"}, "answer": "B", "explanation": "The Delegation of Control Wizard grants precisely-scoped permissions on exactly one OU — least privilege. Account Operators (A) can manage most accounts domain-wide, which is far broader than needed. Full Control on the domain (C) or local admin on DCs (D) effectively hands over the directory. Scoped OU delegation is the exam's expected answer for 'delegate X without making them an admin.'", "domain": null}
{"id": "bb-01508", "exam": "Microsoft Azure Fundamentals", "exam_code": "AZ-900", "product": "Study Guide", "question": "A company deploys an application update to production in minutes instead of waiting weeks for hardware procurement. Which benefit does this demonstrate?", "options": {"A": "Reliability", "B": "Agility", "C": "Predictability", "D": "High availability"}, "answer": "B", "explanation": "Agility is the ability to rapidly provision, change, and deprovision resources — business responsiveness. Predictability is about forecasting; reliability is about recovery.", "domain": "Domain 1"}
{"id": "bb-05333", "exam": "TEAS 7", "exam_code": "TEAS-7", "product": "Practice Pack", "question": "Read the discharge instructions: 'For the first 24 hours, keep the surgical dressing dry and intact. On day two, you may shower but do not scrub the incision; pat it dry afterward. If redness spreads more than an inch from the incision, call the clinic the same day. At your one-week follow-up, the surgeon will decide whether the sutures can be removed.' Which action comes SECOND in the sequence?", "options": {"A": "Calling the clinic about spreading redness.", "B": "Having the sutures removed at the follow-up.", "C": "Showering on day two without scrubbing the incision.", "D": "Keeping the dressing dry for the first 24 hours."}, "answer": "C", "explanation": "C is correct because the instructions follow a timeline: first keep the dressing dry for 24 hours, then (second) shower on day two without scrubbing the incision. A is wrong because calling about spreading redness is conditional ('if'), not a scheduled second step in the sequence. B is wrong because suture removal happens at the one-week follow-up, which is the last event in the sequence. D is wrong because keeping the dressing dry is explicitly the first 24-hour instruction — the first step, not the second.", "domain": null}
{"id": "bb-00024", "exam": "Google Professional ML Engineer", "exam_code": null, "product": "Practice Pack", "question": "A hospital network deploys a patient readmission-risk model on Vertex AI. HIPAA policy requires that prediction traffic stay inside the hospital's VPC — no public internet path may exist. Which configuration meets the requirement?", "options": {"A": "A Vertex AI private endpoint using Private Service Connect, with public access disabled", "B": "A standard Vertex AI public endpoint protected by an API key in the request header", "C": "Deploying the model to Cloud Run and restricting ingress to internal traffic", "D": "Putting the public endpoint behind Cloud CDN with signed URLs"}, "answer": "A", "explanation": "A is correct because Vertex AI private endpoints expose the model through Private Service Connect inside the customer's VPC with no public IP — the request never traverses the internet, satisfying the hard isolation requirement. B is wrong: a public endpoint with an API key is still reachable from the internet; key-based auth does not remove network exposure. C is wrong: the requirement is to serve the Vertex AI model, and moving it to Cloud Run changes the serving stack without providing the private-endpoint guarantee for the Vertex deployment. D is wrong: a CDN with signed URLs is still a public internet surface; signed URLs control who can fetch, not where traffic flows.", "domain": null}
{"id": "bb-03628", "exam": "CompTIA Network+", "exam_code": "N10-009", "product": "Practice Pack", "question": "A user opens a website and data travels from the browser down through the OSI layers, each layer adding its own header. This process is called:", "options": {"A": "Decapsulation", "B": "Encapsulation", "C": "Segmentation", "D": "Fragmentation"}, "answer": "B", "explanation": "Encapsulation = each layer adds its header as data moves DOWN the stack. Decapsulation is the reverse on the receiving end. CompTIA loves this pair - don't mix them up. ByteBar · Network+ N10-009 Practice Pack — for personal use only 2", "domain": null}
{"id": "bb-00725", "exam": "Google Associate Cloud Engineer", "exam_code": "ACE", "product": "Practice Pack", "question": "A contractor should be able to read objects in a bucket only during business hours. How can this time restriction be enforced?", "options": {"A": "A VPC firewall rule", "B": "A billing budget", "C": "An IAM Condition with a time restriction on the role binding", "D": "A separate project for the contractor"}, "answer": "C", "explanation": "IAM Conditions add attributes such as time windows to a binding, so access is only valid during business hours. Firewall rules filter network traffic, not IAM permissions.", "domain": "Section 4"}
{"id": "bb-01145", "exam": "Microsoft Azure Administrator", "exam_code": "AZ-104", "product": "Study Guide", "question": "Security policy requires all internet-bound traffic from Azure VMs to be inspected by the on-premises security stack. The VNets connect to on-premises via ExpressRoute. What should you configure?", "options": {"A": "A NAT gateway on each subnet", "B": "Forced tunneling by advertising a default route (0.0.0.0/0) from the on-premises router over the ExpressRoute BGP session", "C": "An Azure Firewall with threat intelligence in each VNet", "D": "Service endpoints for all Azure PaaS traffic"}, "answer": "B", "explanation": "Forced tunneling advertises a 0.0.0.0/0 route from on-premises over the ExpressRoute BGP session, so internet-bound traffic hairpins through the on-premises security stack for inspection. A NAT gateway (A) sends traffic directly to the internet — the opposite of the requirement; Azure Firewall (C) inspects in Azure, not on-premises.", "domain": null}
{"id": "bb-04869", "exam": "Microsoft SC-500", "exam_code": "SC-500", "product": "Study Guide", "question": "Tailwind Traders needs to run event-driven containerized microservices that scale to zero, with built-in Dapr and KEDA, without managing Kubernetes nodes. Which Azure service should they choose?", "options": {"A": "Azure Container Instances", "B": "Azure Container Apps", "C": "AKS with cluster autoscaler", "D": "Azure Service Fabric"}, "answer": "B", "explanation": "Azure Container Apps is the serverless container platform: it scales to zero on events, integrates Dapr and KEDA natively, and removes node and orchestrator management. Container Instances runs containers but lacks scale-to-zero, Dapr, and revisions. AKS always carries cluster management overhead even with autoscaling. Service Fabric is a distributed-systems platform with a heavier operational model, not serverless scale-to-zero containers.", "domain": null}
{"id": "bb-03137", "exam": "CompTIA SecAI+", "exam_code": "CY0-001", "product": "Study Guide", "question": "A vendor claims its deep-learning IDS detects novel attacks because its stacked layers learn useful features directly from raw packet data, with no manual feature design. What architectural idea makes this possible?", "options": {"A": "Decision trees, which split traffic on hand-picked thresholds", "B": "K-means clustering, which groups packets without learning features", "C": "Neural networks, whose layered neurons automatically learn hierarchical representations from raw input", "D": "Bayesian filters, which apply fixed spam rules to packets"}, "answer": "C", "explanation": "Neural networks is correct because deep learning stacks layers of neurons that automatically learn increasingly abstract representations, from raw bytes up to attack behaviors, which is why no manual feature engineering is needed. Decision trees is wrong because trees split on features a human provides; they do not learn representations from raw data. K-means clustering is wrong because it merely groups similar points; it does not learn hierarchical features. Bayesian filters is wrong because they apply probabilistic rules to predefined features rather than discovering features themselves.", "domain": null}
{"id": "bb-03626", "exam": "CompTIA Network+", "exam_code": "N10-009", "product": "Practice Pack", "question": "A user can ping 8.8.8.8 but can't open google.com in a browser. The problem is MOST likely:", "options": {"A": "DNS resolution", "B": "The Ethernet cable", "C": "DHCP", "D": "The default gateway"}, "answer": "A", "explanation": "IP works + names fail = DNS. The gateway and cable are proven good by the successful ping. Check DNS server settings and try nslookup next.", "domain": null}
{"id": "bb-02546", "exam": "Certified Kubernetes Application Developer", "exam_code": "CKAD", "product": "Study Guide", "question": "kubectl expose deploy web --port=8080 --target-port=80 creates a Service, but traffic to the Service fails. The container listens on port 80 and clients call the Service on 8080. What should you check first?", "options": {"A": "The port mapping is wrong — swap the flags to --port=80 --target-port=8080 ByteBar · Certified Kubernetes Application Developer (CKAD) Study Guide — for personal use only 53", "B": "Whether the Service has endpoints: kubectl get endpoints web — a selector/label mismatch is the most common cause when the port mapping is correct", "C": "--target-port must reference a named port, not a number — rename the container port", "D": "The Service type must be NodePort for port mappings to work"}, "answer": "B", "explanation": "The mapping is actually correct as written (Service:8080 → container:80). When the mapping is right but traffic fails, check endpoints first — an empty endpoint list means the selector matches no pods, and that's the most common cause. Don't \"fix\" correct YAML; verify the data path. Swapping the flags (A) would break a working mapping.", "domain": null}
{"id": "bb-01063", "exam": "AWS Certified AI Practitioner", "exam_code": "AIF-C01", "product": "Study Guide", "question": "A team trains a custom model with full control of the pipeline and deploys it to autoscaling GPU endpoints with experiment tracking. Which service fits?", "options": {"A": "Amazon Bedrock", "B": "Amazon SageMaker AI", "C": "Amazon Rekognition", "D": "Amazon Q Business"}, "answer": "B", "explanation": "SageMaker AI is the full ML platform - training, tuning, and deploying custom models at scale - when you need control beyond Bedrock's managed models. BYTEBAR · SWEET IT SKILLS, NO FLUFF ByteBar | AWS Certified AI Practitioner (AIF-C01) Study Guide", "domain": "Domain 3"}
{"id": "bb-01664", "exam": "ISC2 Certified in Cybersecurity", "exam_code": "CC", "product": "Practice Pack", "question": "Bonuses.\" This technique is:", "options": {"A": "Pretexting", "B": "Baiting", "C": "Vishing", "D": "Whaling"}, "answer": "B", "explanation": "Baiting = dangling something tempting (infected media, fake download) so the victim infects themselves. Pretexting invents a scenario; vishing uses voice calls; whaling targets executives.", "domain": null}
{"id": "bb-01254", "exam": "Microsoft AZ-802", "exam_code": "AZ-802", "product": "Practice Pack", "question": "Fabrikam is decommissioning an on-premises file server that is onboarded to Azure Arc. The server will be moved to an isolated lab with no Azure connectivity, but the Connected Machine agent software must stay installed so it can be reused later. Which command removes the server's Arc registration while leaving the agent installed?", "options": {"A": "azcmagent disconnect", "B": "azcmagent check", "C": "Uninstall the Connected Machine agent from the server", "D": "Delete the machine resource in the Azure portal"}, "answer": "A", "explanation": "azcmagent disconnect — azcmagent disconnect cleanly removes the Arc registration and clears the agent's local connection state while leaving the software installed for future reuse. azcmagent check only tests network connectivity and changes nothing. Uninstalling removes the agent entirely, which the scenario forbids. Deleting the portal resource alone leaves the agent configured against a registration that no longer exists, causing repeated connection failures.", "domain": null}
{"id": "bb-05213", "exam": "CompTIA Security+", "exam_code": "SY0-701", "product": "Study Guide", "question": "A switch feature builds a table of legitimate DHCP server ports and drops DHCP offers arriving on untrusted access ports. Which attack does this defeat?", "options": {"A": "ARP spoofing", "B": "Rogue DHCP", "C": "MAC flooding", "D": "DNS amplification"}, "answer": "B", "explanation": "DHCP snooping trusts only designated ports for DHCP offers — a rogue server plugged into a wall jack gets silenced. ARP spoofing needs dynamic ARP inspection, MAC flooding needs port security, and DNS amplification is an upstream flood. Trust the infrastructure port, not the cubicle port.", "domain": null}
{"id": "bb-00735", "exam": "Google Associate Cloud Engineer", "exam_code": "ACE", "product": "Practice Pack", "question": "A new hire asks for broad administrative access to everything so they can get started quickly. Following Google Cloud best practices, what should the administrator do?", "options": {"A": "Grant the Owner role as requested", "B": "Grant only the predefined roles needed for their specific tasks", "C": "Create a new organization for the hire", "D": "Give them a service account key with Editor access ByteBar · Associate Cloud Engineer Practice Pack — for personal use only 22"}, "answer": "B", "explanation": "Least privilege means starting with the minimal predefined roles for the person's tasks and expanding only as needed. Owner-on-day-one and standing service account keys both violate that principle. ByteBar · Associate Cloud Engineer Practice Pack — for personal use only 34 Score yourself Score What it means Next step 85–100 Exam-ready Book the exam. Light review of misses only. 75–84 Close Re-study your two weakest domains, then retake this pack. 60–74 Gaps Rebuild fundamentals in the domains you missed most, then retake. Below 60 Not yet Study core concepts first and give it two more weeks before retaking. Domain answer map Questions 1–20: Section 1: Setting up a cloud solution environment · Questions 21–50: Section 2: Planning and implementing a cloud solution · Questions 51–80: Section 3: Ensuring the successful operation of a cloud solution · Questions 81–100: Section 4: Configuring access and security EXAM TIP — Your weakest domain by percentage — not by raw misses — is what to study next. 85+ and you're ready. Go earn it. ByteBar · Associate Cloud Engineer Practice Pack — for personal use only 35 BYTEBAR · SWEET IT SKILLS, NO FLUFF A final note ByteBar is an independent study resource and is not affiliated with or endorsed by Google LLC. All 100 questions in this pack are original practice questions written by ByteBar for study purposes. They are not real exam questions, recalled exam content, or exam dumps, and they do not guarantee that you will pass the Associate Cloud Engineer exam. Google Cloud and related marks are trademarks of Google LLC. This product is an independent study aid — it is not sponsored, endorsed, or approved by Google. Good luck — and go earn it. ByteBar · Associate Cloud Engineer Practice Pack — for personal use only 36", "domain": "Section 4"}
{"id": "bb-05081", "exam": "Subnetting Mastery", "exam_code": "SUBNET", "product": "Study Guide", "question": "Your VLSM plan used 10.70.0.0/23, 10.70.2.0/25 and 10.70.2.128/26 out of 10.70.0.0/16. How many addresses remain unused?", "options": {"A": "64,832", "B": "64,896", "C": "64,768", "D": "65,024"}, "answer": "A", "explanation": "Used: 512 + 128 + 64 = 704 addresses; 65,536 - 704 = 64,832 still free in the /16.", "domain": null}
{"id": "bb-02013", "exam": "ISC2 CCSP", "exam_code": "CCSP", "product": "Practice Pack", "question": "Cloud Application Security After a legal review flags copyleft-licensed code inside a commercial SaaS product, the CISO wants automated detection of risky open-source licenses in every build. Which capability provides this?", "options": {"A": "SCA", "B": "DAST", "C": "RASP", "D": "SAST"}, "answer": "A", "explanation": "Software Composition Analysis (A) inventories third-party and open-source components and flags both known CVEs and license risks such as copyleft GPL code in a commercial product. DAST (B) tests running apps for runtime flaws. RASP (C) protects apps at runtime. SAST (D) analyzes first-party source code, not dependency licenses.", "domain": "Cloud Application Security"}
{"id": "bb-04475", "exam": "Red Hat RHCSA/RHCE", "exam_code": "RHCSA", "product": "Practice Pack", "question": "SSH key authentication fails and the server log says 'Authentication refused: bad ownership or modes for /home/bob/.ssh/authorized_keys'. The directory is 755 and the file is 644. What is the fix?", "options": {"A": "Run chmod 700 ~/.ssh and chmod 600 ~/.ssh/authorized_keys, and make sure the home directory itself is not group- or world-writable (sshd's StrictModes requires this)", "B": "Restarting sshd is the only fix; permissions on key files are irrelevant", "C": "Set both to 777 so the sshd process can always read them", "D": "Mode 644 on authorized_keys is correct and secure; the error must come from elsewhere"}, "answer": "A", "explanation": "is not group- or world-writable (sshd's StrictModes requires this) sshd's StrictModes refuses key authentication when ~/.ssh or authorized_keys are writable by others (or the home directory is group-writable), because another user could then plant keys. The secure standard is 700 on ~/.ssh, 600 on authorized_keys, and a non-writable home directory. Option D is wrong: 644 leaves the file world-readable and, combined with a writable path, triggers the refusal. Option C is the opposite of secure. Option B ignores the explicit log message naming ownership/modes.", "domain": "Networking, Users & Security"}
{"id": "bb-00757", "exam": "Google Associate Cloud Engineer", "exam_code": "ACE", "product": "Practice Pack", "question": "A VM is unreachable over SSH after a firewall rule change locked everyone out. How can an administrator regain console access to fix it?", "options": {"A": "Assign the VM a new external IP address", "B": "Delete and recreate the VM", "C": "Enable Cloud NAT", "D": "Connect to the serial console with gcloud compute connect-to-serial-port"}, "answer": "D", "explanation": "The serial console provides out-of-band access independent of SSH and firewall rules. A new IP or NAT does not bypass the blocking firewall rule, and deleting the VM destroys its state. ByteBar · Associate Cloud Engineer Practice Pack — for personal use only 30", "domain": "Section 3"}
{"id": "bb-00558", "exam": "CompTIA A+", "exam_code": "220-1201/220-1202", "product": "Practice Pack", "question": "Two employees need to share a large video file between phones with no Wi-Fi or cellular signal. Which technology works at very short range for tap-to-share?", "options": {"A": "Bluetooth 5.0", "B": "NFC-initiated transfer", "C": "Hotspot tethering", "D": "5G ByteBar · CompTIA A+ Practice Pack — for personal use only 3"}, "answer": "B", "explanation": "NFC works at a few centimeters — ideal for tap-to-pair/initiating a transfer. Bluetooth then typically carries the actual file.", "domain": null}
{"id": "bb-01943", "exam": "Adult CCRN", "exam_code": "CCRN", "product": "Study Guide", "question": "A patient with chronic kidney disease and diabetes is scheduled for a contrast CT. The most evidence-based nursing intervention to prevent contrast-induced acute kidney injury is:", "options": {"A": "Restricting fluids to avoid overload", "B": "Pre- and post-procedure isotonic IV hydration per protocol", "C": "N-acetylcysteine mega-doses", "D": "High-dose loop diuretics before contrast"}, "answer": "B", "explanation": "Isotonic IV hydration before and after contrast is the proven kidney protection — it dilutes the contrast and maintains tubular flow. N-acetylcysteine has weak, inconsistent evidence; restricting fluids concentrates the contrast in the tubules; diuretics volume-deplete the patient and worsen injury.", "domain": null}
{"id": "bb-04050", "exam": "PMHNP-BC", "exam_code": "PMHNP-BC", "product": "Practice Pack", "question": "After an amitriptyline overdose, a patient is confused, flushed, febrile, with dry skin, dilated pupils, urinary retention, and absent bowel sounds. Which toxidrome is this?", "options": {"A": "Cholinergic toxidrome with SLUDGE symptoms and pinpoint pupils.", "B": "Serotonin syndrome with clonus, hyperreflexia, and diaphoresis.", "C": "Anticholinergic toxidrome: 'blind as a bat, hot as a hare, dry as a bone, mad as a hatter.'", "D": "Neuroleptic malignant syndrome with lead-pipe rigidity and bradykinesia."}, "answer": "C", "explanation": "Tricyclic antidepressants are potent anticholinergics in overdose, producing the classic anticholinergic toxidrome: mydriasis causing blurred vision, hyperthermia from impaired sweating, dry flushed skin and mucous membranes, urinary retention, ileus, tachycardia, and delirium, memorized as blind, hot, red, dry, and mad. Recognizing the toxidrome directs care toward supportive management and physostigmine in severe cases rather than treatments for other syndromes. A cholinergic toxidrome looks opposite, with salivation, lacrimation, urination, diarrhea, and pinpoint pupils. Serotonin syndrome features clonus, hyperreflexia, and sweaty, clammy skin rather than dry heat. Neuroleptic malignant syndrome centers on lead-pipe rigidity and bradykinesia from dopamine blockade, not the anticholinergic constellation seen here.", "domain": null}
{"id": "bb-00179", "exam": "Google Professional ML Engineer", "exam_code": null, "product": "Study Guide", "question": "An automated retraining pipeline produces a new model weekly. How should promotion to production be gated?", "options": {"A": "Immediate 100% rollout of the retrained model", "B": "No testing — the retraining pipeline is trusted", "C": "A canary analysis comparing error rates before full promotion", "D": "Testing only on training data"}, "answer": "C", "explanation": "Even automated retraining needs a promotion gate: canary the candidate against live traffic and compare error/business metrics before full rollout. 100% rollout (A) risks a bad model hitting all users. Blind trust (B) is how regressions ship. Training-data testing (D) cannot validate production behavior. Retraining automation still needs canary promotion.", "domain": null}
{"id": "bb-05557", "exam": "CompTIA Linux+", "exam_code": "XK0-006", "product": "Practice Pack", "question": "Which command shows which process is using TCP port 443?", "options": {"A": "lsof -i :443", "B": "ps aux | grep 443", "C": "ls -l /proc/443", "D": "df -h | grep 443"}, "answer": "A", "explanation": "lsof -i :443 lists open files (sockets) matching port 443, revealing the owning process. ps aux | grep 443 only finds processes with '443' in their command line, /proc/443 would be a PID directory (not a port), and df reports disk space.", "domain": "Domain 2"}
{"id": "bb-03530", "exam": "ITIL 4 Foundation", "exam_code": "ITIL-4", "product": "Practice Pack", "question": "Key concepts Which statement BEST describes value co-creation?", "options": {"A": "The provider creates all value; the consumer only pays for it", "B": "Value emerges from active collaboration between providers and consumers — the provider facilitates outcomes, but the consumer's participation determines value", "C": "Value is created only during the sales process", "D": "Consumers create value independently of providers"}, "answer": "B", "explanation": "outcomes, but the consumer's participation determines value Providers cannot deliver value unilaterally — they co-create it with consumers through service relationships. The consumer's use of the service is what realizes the outcome and the value.", "domain": "Key concepts"}
{"id": "bb-04105", "exam": "PMHNP-BC", "exam_code": "PMHNP-BC", "product": "Study Guide", "question": "A patient develops delusions and disorganized behavior 10 days ago, with full return to baseline expected. No mood episode. What is the most likely diagnosis?", "options": {"A": "Schizophreniform disorder", "B": "Schizophrenia", "C": "Brief psychotic disorder", "D": "Delusional disorder"}, "answer": "C", "explanation": "Brief psychotic disorder = 1+ day and less than 1 month, with eventual FULL return to premorbid functioning. Schizophreniform (A) is 1-6 months. Schizophrenia (B) needs 6 months. Delusional disorder (D) is nonbizarre delusions for 1+ month with otherwise intact functioning. Duration is the entire differential here.", "domain": null}
{"id": "bb-03883", "exam": "NCLEX-RN", "exam_code": "NCLEX-RN", "product": "Practice Pack", "question": "Which client is at highest risk for a fall?", "options": {"A": "A 45-year-old on postoperative day 1 ambulating with a steady gait.", "B": "A 78-year-old taking a diuretic and an antihypertensive who reports dizziness when standing.", "C": "A 30-year-old with a sprained ankle using crutches with a correct gait.", "D": "A 60-year-old with well-controlled diabetes and normal vision."}, "answer": "B", "explanation": "The 78-year-old has multiple compounding fall risk factors: advanced age, a diuretic and an antihypertensive that cause orthostatic hypotension, and reported dizziness on standing. Option A is lower risk because the gait is steady. Option C is lower risk because the client uses crutches correctly. Option D is lower risk with controlled diabetes and normal vision. Fall prevention targets the client with the most risk factors, so the nurse prioritizes interventions like a bed alarm and supervised ambulation for the older adult.", "domain": null}
{"id": "bb-03585", "exam": "ITIL 4 Foundation", "exam_code": "ITIL-4", "product": "Practice Pack", "question": "Who ultimately determines whether a service delivers value?", "options": {"A": "The service provider's management team", "B": "The change advisory board", "C": "The authors of the ITIL framework", "D": "The service consumer, based on their own perception of the outcomes"}, "answer": "D", "explanation": "Value is subjective and defined by the consumer's perception — the provider facilitates outcomes, but the consumer judges whether value was delivered. Providers cannot declare value on the consumer's behalf.", "domain": "Guiding principles (7)"}
{"id": "bb-01607", "exam": "ISC2 Certified in Cybersecurity", "exam_code": "CC", "product": "Practice Pack", "question": "A company document states: \"All remote access must use MFA.\" A second document lists the exact steps to enroll in the MFA app. The second document is a:", "options": {"A": "Policy", "B": "Standard", "C": "Procedure", "D": "Regulation"}, "answer": "C", "explanation": "Step-by-step how-to instructions = procedure. The \"must use MFA\" mandate is the policy; a standard would give mandatory specifics (e.g., which MFA methods are approved).", "domain": null}
{"id": "bb-05247", "exam": "CompTIA Security+", "exam_code": "SY0-701", "product": "Study Guide", "question": "Before analyzing a disk image, the analyst records its SHA-256 hash. Months later in court, the defense claims the image was tampered with. How does the analyst refute this?", "options": {"A": "By showing the write blocker logs", "B": "By re-imaging the original drive", "C": "By recomputing the hash and showing it matches", "D": "By presenting the analyst's credentials"}, "answer": "C", "explanation": "Recomputing the hash proves the image is bit-for-bit identical — any alteration changes the hash. Write blockers prevent writes during acquisition, re-imaging may be impossible, and credentials don't prove integrity. Hash at acquisition, hash again whenever challenged.", "domain": null}
{"id": "bb-04550", "exam": "Red Hat RHCSA/RHCE", "exam_code": "RHCSA", "product": "Study Guide", "question": "You need firewalld to allow HTTPS permanently on webservers and ensure firewalld is running. Which tasks are correct?", "options": {"A": "- name: Allow HTTPS ansible.builtin.iptables: chain: INPUT protocol: tcp destination_port: 443 jump: ACCEPT", "B": "- name: Allow HTTPS ansible.posix.firewalld: service: https state: enabled notify: reload firewalld", "C": "- name: Allow HTTPS ansible.posix.firewalld: service: https permanent: true state: enabled immediate: true - name: Ensure firewalld running ansible.builtin.service: name: firewalld state: started enabled: true", "D": "- name: Allow HTTPS ansible.builtin.firewalld: port: 443/tcp permanent: yes state: present - name: Ensure firewalld running ansible.builtin.systemd: name: firewalld state: started"}, "answer": "C", "explanation": "C is correct: the firewalld module lives in the ansible.posix collection (fully qualified ansible.posix.firewalld), uses service: https with permanent: true and state: enabled, immediate: true applies it to the runtime config too, and the service module ensures firewalld itself runs. D is wrong: there is no ansible.builtin.firewalld module and 'state: present' is not valid for it. A is wrong: iptables rules bypass firewalld and will not survive a firewalld reload; also the task never ensures firewalld is running. B is incomplete: without permanent: true the rule is runtime-only and lost on reload, and notify: requires a matching handler to exist.", "domain": null}
{"id": "bb-02353", "exam": "Certified Kubernetes Administrator", "exam_code": "CKA", "product": "Study Guide", "question": "Control-plane nodes carry the taint A monitoring DaemonSet must node-role.kubernetes.io/control-plane:NoSchedule. still run on them. How?", "options": {"A": "Remove the taint from the control-plane nodes", "B": "Add a matching toleration for with effect node-role.kubernetes.io/control-plane NoSchedule to the DaemonSet's pod template", "C": "Set on the DaemonSet — host networking bypasses taints hostNetwork: true", "D": "Give the DaemonSet — priority overrides taints priorityClassName: system-node-critical"}, "answer": "B", "explanation": "Tolerations are the designed mechanism: the DaemonSet tolerates the control-plane taint. Removing the taint invites regular workloads onto control planes, hostNetwork does not bypass scheduling, and priority affects preemption, not taints.", "domain": null}
{"id": "bb-04684", "exam": "Microsoft SC-500", "exam_code": "SC-500", "product": "Practice Pack", "question": "A Contoso contractor requests PIM activation of the Contributor role for 8 hours, but the activation fails. The administrator confirms the role settings allow a maximum activation duration of 4 hours. What should be changed to allow the 8-hour activation?", "options": {"A": "Convert the contractor's eligible assignment to an active permanent assignment", "B": "Have the contractor submit two consecutive 4-hour activation requests", "C": "Upgrade the contractor's license from Microsoft Entra ID P1 to P2", "D": "Update the PIM role settings for the Contributor role to allow a maximum activation duration of 8 hours"}, "answer": "D", "explanation": "Update the PIM role settings for the Contributor role to allow a maximum activation duration of 8 hours — The failure comes from the role settings, not the request: the maximum activation duration configured for the Contributor role caps every activation at 4 hours. Raising that setting to 8 hours fixes the root cause for all future requests. A abandons just-in-time access entirely. B is a clumsy workaround that still violates the spirit of the 4-hour cap. C is irrelevant because PIM activation duration is governed by role settings, not license tier.", "domain": null}
{"id": "bb-03284", "exam": "Docker Certified Associate", "exam_code": "DCA", "product": "Practice Pack", "question": "Your company runs a metrics collector agent that must run on every single node in a 12-node Swarm — including any nodes added next month. Which service mode, and why?", "options": {"A": "Replicated with a placement constraint of node.role == worker, which covers all nodes", "B": "A stack with deploy.replicas: 12, which behaves like global mode", "C": "Replicated with --replicas 12, because it guarantees exactly 12 tasks", "D": "Global mode (--mode global), because Swarm places exactly one task per node automatically, including future nodes ByteBar · DCA Practice Pack — for personal use only 3"}, "answer": "D", "explanation": "future nodes Global mode is purpose-built for 'one task on every node' — log collectors, monitoring agents, security daemons. It automatically covers nodes added later, which replicated --replicas 12 does NOT (you'd have to rescale). Constraints don't create per-node placement either; they only filter eligible nodes. EXAM TRAP: replicated vs global is the single most-tested service concept — replicated = N tasks spread across nodes (web apps), global = exactly one per node (agents). ByteBar · DCA Practice Pack — for personal use only 30", "domain": "Domain 1"}
{"id": "bb-01854", "exam": "Adult CCRN", "exam_code": "CCRN", "product": "Practice Pack", "question": "A patient with syndrome of inappropriate antidiuretic hormone (SIADH) has sodium 128 mEq/L, is euvolemic, and is asymptomatic. The nurse anticipates first-line management of:", "options": {"A": "Hypotonic IV fluids", "B": "Desmopressin", "C": "Fluid restriction, reserving hypertonic saline for severe neurologic symptoms such as seizures or coma", "D": "Loop diuretics as first-line therapy"}, "answer": "C", "explanation": "SIADH produces euvolemic hyponatremia from excess ADH-mediated water retention; first-line therapy is fluid restriction, often to 800 to 1000 mL per day, addressing the water excess directly. Severe neurologic symptoms demand hypertonic 3 percent saline to raise sodium acutely, with chronic correction limited to about 8 to 10 mEq/L per 24 hours to avoid osmotic demyelination. Hypotonic fluids worsen the hyponatremia. Desmopressin treats diabetes insipidus, the opposite disorder, and would deepen SIADH hyponatremia.", "domain": null}
{"id": "bb-01564", "exam": "Microsoft Azure Fundamentals", "exam_code": "AZ-900", "product": "Study Guide", "question": "[Domain 1] — A hospital must keep patient records in its own datacenter for regulatory reasons but wants to burst extra compute capacity to Azure during seasonal peaks. Which deployment model fits best?", "options": {"A": "Public cloud only", "B": "Private cloud only", "C": "Hybrid cloud", "D": "On-premises only"}, "answer": "C", "explanation": "Hybrid cloud combines public and private clouds with shared data and apps - sensitive workloads stay on-premises while elastic demand bursts to the public cloud.", "domain": null}
{"id": "bb-01822", "exam": "Adult CCRN", "exam_code": "CCRN", "product": "Practice Pack", "question": "A patient presents with fever, right upper quadrant pain, and jaundice, known as Charcot triad. Hours later she becomes hypotensive and confused, completing Reynolds pentad. The nurse recognizes:", "options": {"A": "Uncomplicated cholelithiasis", "B": "Ascending cholangitis with biliary septic shock, anticipating emergent biliary drainage via ERCP, broad-spectrum antibiotics, and sepsis resuscitation", "C": "Viral hepatitis", "D": "Pancreatitis alone"}, "answer": "B", "explanation": "Charcot triad of fever, right upper quadrant pain, and jaundice indicates ascending cholangitis from biliary obstruction and infection; adding hypotension and altered mentation, Reynolds pentad, marks severe disease with septic shock. This is a medical emergency requiring urgent biliary decompression, usually ERCP with stent or stone extraction, plus broad-spectrum antibiotics and full sepsis resuscitation. Delay risks rapid death. Uncomplicated gallstones lack the infectious signs, viral hepatitis lacks this pain pattern, and pancreatitis presents differently. Section 11 — Renal II", "domain": null}
{"id": "bb-03344", "exam": "Family Nurse Practitioner", "exam_code": "FNP", "product": "Practice Pack", "question": "A 58-year-old man with hypertension has office readings of 148/92 mmHg but reports home readings averaging 128/80 mmHg on a validated monitor with correct technique. The best interpretation and next step is:", "options": {"A": "Ignore the home readings — office readings are the only valid measure", "B": "Likely white-coat hypertension — confirm with home or ambulatory monitoring averages before intensifying therapy", "C": "Stop all antihypertensives immediately", "D": "Send him to the emergency department"}, "answer": "B", "explanation": "Office readings substantially higher than properly obtained home averages suggest white-coat hypertension, and guidelines require confirming with out-of-office measurements — averaged home readings or 24-hour ambulatory monitoring — before intensifying therapy, since treating office numbers alone causes overtreatment and hypotension. Dismissing home readings discards the more prognostically meaningful data. Stopping antihypertensives based on one discrepancy is reckless, and the emergency department is for hypertensive emergency, not an asymptomatic office-office gap. Verify his home technique and device validation first; if out-of-office averages confirm control, continue current therapy with periodic out-of-office surveillance.", "domain": null}
{"id": "bb-04281", "exam": "CompTIA PenTest+", "exam_code": "PT0-003", "product": "Practice Pack", "question": "Your Nessus scan and your OpenVAS scan report different severity counts for the same subnet. What is the most likely reason?", "options": {"A": "Different plugin feeds and detection logic between the scanners", "B": "One of the scanners is always wrong", "C": "CVSS scores change every 24 hours", "D": "Scanners cannot scan the same subnet twice"}, "answer": "A", "explanation": "Nessus and OpenVAS are built by different vendors with different plugin feeds, detection signatures, and coverage priorities, so their finding counts routinely disagree — neither is automatically authoritative. Declaring one always wrong skips the professional step of manually validating contested findings. CVSS base scores are stable reference values, not daily fluctuations. And rescanning the same subnet is standard practice for tracking remediation; the disagreement comes from the tools' logic, not from any prohibition on repeat scans.", "domain": null}
{"id": "bb-04870", "exam": "Microsoft SC-500", "exam_code": "SC-500", "product": "Study Guide", "question": "Tailwind Traders requires that a storage account be reachable only through private endpoints, with no public internet path whatsoever — even for callers presenting valid credentials. Which setting enforces this?", "options": {"A": "Set 'Public network access' to Disabled on the storage account.", "B": "Set the firewall default action to Deny.", "C": "Enable 'Allow trusted Microsoft services to access this storage account.'", "D": "Disable shared key access on the storage account."}, "answer": "A", "explanation": "Setting public network access to Disabled removes the public endpoint entirely, so all requests must arrive through private endpoints regardless of firewall rules or credentials. A Deny-default firewall (B) still permits explicitly allowed rules and the trusted-services exception. Option C is an exception that widens access, the opposite of the requirement. Disabling shared key access (D) changes the permitted authentication methods but leaves the public network path intact.", "domain": null}
{"id": "bb-04370", "exam": "CompTIA PenTest+", "exam_code": "PT0-003", "product": "Study Guide", "question": "You need to build a target list of employee email addresses for a later phishing simulation, using only public information. Which tool is designed for this OSINT task?", "options": {"A": "Search breach data and public sources for employee emails and names", "B": "Scan the target's internal network for SMB shares", "C": "Exploit an SQL injection vulnerability", "D": "Query the target's mail server for user accounts"}, "answer": "A", "explanation": "theHarvester aggregates emails, names, subdomains, and hosts from public sources (search engines, PGP key servers, breach data) — pure OSINT. Querying the mail server via SMTP VRFY/EXPN is active enumeration. SMB share scanning is active network enumeration. Exploiting SQLi is an attack, not reconnaissance.", "domain": null}
{"id": "bb-03627", "exam": "CompTIA Network+", "exam_code": "N10-009", "product": "Practice Pack", "question": "A user can't reach any network resource. Per the CompTIA troubleshooting methodology, the FIRST step is:", "options": {"A": "Implement a solution", "B": "Identify the problem - gather information and question users", "C": "Test the theory", "D": "Document the findings"}, "answer": "B", "explanation": "Methodology order: identify the problem, establish a theory, test it, plan, implement, verify, document. Always START by defining the problem - jumping to fixes wastes time.", "domain": null}
{"id": "bb-01870", "exam": "Adult CCRN", "exam_code": "CCRN", "product": "Practice Pack", "question": "Compared with a standard advance directive, a POLST (Physician Orders for Life-Sustaining Treatment) form:", "options": {"A": "Is intended only for healthy young adults", "B": "Carries no medical authority", "C": "Translates goals of care into actionable medical orders that travel with the patient across care settings", "D": "Cannot address resuscitation preferences"}, "answer": "C", "explanation": "POLST, known as MOLST or POST in some states, is designed for seriously ill patients: it converts treatment preferences about CPR, intubation, artificial nutrition, and hospitalization into signed, immediately actionable medical orders honored across hospitals, nursing facilities, and emergency medical services. Advance directives guide future decisions when capacity is lost; POLST provides orders for the present. Both require informed discussion with the patient or surrogate and periodic review as conditions change.", "domain": null}
{"id": "bb-03222", "exam": "Docker Certified Associate", "exam_code": "DCA", "product": "Practice Pack", "question": "A monitoring task must run on every node, but the team also wants exactly 3 copies of the web API spread across workers, with the API never landing on managers. Which combination is correct?", "options": {"A": "One stack with deploy.mode: global for both services", "B": "Two services: agent with --mode global; web with --replicas 3 --constraint 'node.role == worker'", "C": "Two services: agent with --replicas ; web with --mode global --constraint 'node.role != manager'", "D": "One service with --mode global --replicas 3"}, "answer": "B", "explanation": "Two different placement needs = two services. The agent is textbook global mode (one task per node, automatic on new nodes). The web API is replicated (3 tasks) with a hard constraint node.role == worker keeping it off managers. --mode global --replicas 3 is contradictory (global ignores replica counts). Replicas set to today's node count won't cover nodes added tomorrow. Constraints use == / != with node.role, node.hostname, node.labels.*, engine.labels.*, node.platform.os/arch.", "domain": "Domain 1"}
{"id": "bb-03390", "exam": "Family Nurse Practitioner", "exam_code": "FNP", "product": "Study Guide", "question": "A 16-year-old requests confidential contraception counseling. The appropriate approach is:", "options": {"A": "Refuse without parental consent", "B": "Provide no counseling", "C": "Explain confidentiality and its limits, then provide confidential counseling", "D": "Notify parents immediately"}, "answer": "C", "explanation": "Explain the limits of confidentiality (harm to self/others, abuse), then protect adolescent privacy — confidentiality drives honest disclosure and care engagement, per state law. Test-Day Playbook • AANPCB: 150 questions, 3 hours — about 72 seconds per question. Flag and return; unanswered questions hurt you. • ANCC: 175 questions, 3.5 hours — same ~72-second pace. Practice the alternate formats (select-all-that-apply, drag-and-drop, hot spot) so format never surprises you. • Read the last line first: the question stem tells you what is actually being asked. Then read the vignette for age, key symptoms, and red flags. • When two answers are both guideline-concordant, choose the safest, least invasive, most cost-effective first step. • \"Most appropriate next step\" in primary care is usually: focused exam or first-line test → first-line therapy → referral only when indicated. • Never choose \"reassure and discharge\" when red flags are present (chest pain, postmenopausal bleeding, thunderclap headache, tearing back pain, suicidal ideation). • Mark questions you're unsure about and move on — momentum beats perfection. • Sleep, eat, arrive early. You've done the work. Disclaimer This guide is a study supplement only. It is not affiliated with, endorsed by, or sponsored by the American Academy of Nurse Practitioners Certification Board (AANPCB) or the American Nurses Credentialing Center (ANCC). Clinical guidelines evolve — verify all management decisions against current guidelines and your supervising resources before applying them in practice. This guide does not replace formal FNP education or clinical training.", "domain": null}
{"id": "bb-01339", "exam": "Microsoft AZ-802", "exam_code": "AZ-802", "product": "Study Guide", "question": "You are a hosting provider running untrusted customer code in Windows containers on shared hosts. Which container isolation mode should you require, and why?", "options": {"A": "Hyper-V isolation, because it runs the container in a lightweight utility VM with its own kernel, isolating it from the host and other tenants", "B": "Process isolation, because it shares the host kernel and is therefore more secure", "C": "Running the container on the host without any isolation", "D": "Hyper-V isolation, because it allows containers to share GPU drivers directly"}, "answer": "A", "explanation": "Windows containers have two isolation modes: process isolation shares the host's kernel (fast and dense, but requires the container OS version to match the host) and Hyper-V isolation runs each container inside a lightweight, optimized utility VM with its own kernel. For untrusted tenant code, Hyper-V isolation is the right choice — a kernel exploit in the container cannot reach the host or neighboring containers. Process isolation (B) is explicitly the less-isolated option. No isolation (C) is not a container at all. GPU sharing (D) is a discrete-device-assignment topic, not a reason to choose Hyper-V isolation.", "domain": null}
{"id": "bb-01977", "exam": "Adult CCRN", "exam_code": "CCRN", "product": "Study Guide", "question": "During handoff, the night nurse gives a rushed, informal report in the hallway. The evidence-based handoff standard the nurse should use is:", "options": {"A": "Texting the report from home later", "B": "Structured handoff (SBAR or I-PASS) with opportunity for questions and read-back of critical items", "C": "Hallway updates are fine for stable patients", "D": "Skipping handoff when the chart is up to date"}, "answer": "B", "explanation": "Handoffs are a top source of harm — structured tools (SBAR/I-PASS), face-to-face exchange, questions, and read-back of critical orders are the standard. Hallway reports lose information; the chart never replaces interactive handoff; delayed texting abandons accountability.", "domain": null}
{"id": "bb-02856", "exam": "CompTIA CySA+", "exam_code": "CS0-004", "product": "Practice Pack", "question": "Which log would BEST help you determine the initial infection vector of a malware outbreak?", "options": {"A": "Email gateway and web proxy logs from before the first detection, showing the malicious message or download", "B": "Printer logs", "C": "Badge access logs for the parking garage", "D": "The CEO's calendar"}, "answer": "A", "explanation": "download The infection vector (phishing email, malicious download) appears in email gateway and proxy logs timestamped before the first host detection. Printer, badge, and calendar data do not show how malware entered.", "domain": "Domain 1"}
{"id": "bb-03615", "exam": "CompTIA Network+", "exam_code": "N10-009", "product": "Practice Pack", "question": "A server is reachable by IP but its hostname won't resolve after a DNS migration. The technician should:", "options": {"A": "Check the DNS zone records, then flush the local resolver cache", "B": "Replace the NIC", "C": "Reinstall the OS", "D": "Change the subnet mask"}, "answer": "A", "explanation": "Stale or missing records after migration = check the zones, then ipconfig /flushdns on the client. Don't rebuild what a cache clear fixes - work from the most likely cause outward.", "domain": null}
{"id": "bb-02415", "exam": "Certified Kubernetes Application Developer", "exam_code": "CKAD", "product": "Practice Pack", "question": "A pod is in CrashLoopBackOff. kubectl logs web-abc shows the CURRENT (just-started, still-booting) container's output, which reveals nothing. How do you see the logs from the container instance that actually crashed?", "options": {"A": "kubectl logs web-abc --all — it aggregates all instances", "B": "kubectl logs web-abc --previous — it shows the last terminated container instance's logs", "C": "Crashed container logs are lost; check the events instead", "D": "Wait for the next crash and run logs faster"}, "answer": "B", "explanation": "--previous (-p) is the CrashLoopBackOff essential: it fetches logs from the previous terminated instance — the one that holds the actual error — instead of the freshly restarted one. Pair it with -c for multi-container pods. The standard crash debug sequence: describe (exit code, events) → logs --previous (the error) → fix.", "domain": "Domain 5"}
{"id": "bb-01608", "exam": "ISC2 Certified in Cybersecurity", "exam_code": "CC", "product": "Practice Pack", "question": "A company replicates its database to a secondary site every 15 minutes. If the primary site fails, at most 15 minutes of transactions can be lost. The 15-minute figure is the:", "options": {"A": "RTO", "B": "RPO", "C": "MTD", "D": "WRT — = tolerable loss, which drives backup/replication"}, "answer": "B", "explanation": "RPO (Recovery Point Objective) data frequency. RTO would be the time allowed to restore service. Data-loss tolerance is always RPO.", "domain": null}
{"id": "bb-00699", "exam": "CompTIA A+", "exam_code": "220-1201/220-1202", "product": "Study Guide", "question": "You need 10 Gbps over 100 meters of copper. Which cable do you choose?", "options": {"A": "Cat5e", "B": "Cat6", "C": "Cat6a", "D": "RG-6"}, "answer": "C", "explanation": "Cat6a supports 10 Gbps to 100 m. Cat6 only reaches 10 Gbps to 55 m; Cat5e tops out at 1 Gbps.", "domain": null}
{"id": "bb-02205", "exam": "Certified Kubernetes Administrator", "exam_code": "CKA", "product": "Practice Pack", "question": "A Deployment with 3 replicas mounts a PVC with accessModes: [ReadWriteOnce]. Two pods schedule to node-a and one to node-b. The node-b pod stays in ContainerCreating with a failed mount event. What is the root cause, and what is the correct fix?", "options": {"A": "ReadWriteOnce allows only ONE node to mount the volume read-write at a time; the node-b pod can never attach it. Fix: use ReadWriteMany (with a supporting provisioner) or run the workload as a StatefulSet with volumeClaimTemplates giving each pod its own volume", "B": "The PVC is too small — increase the storage request", "C": "The StorageClass provisioner is down — restart the external-provisioner pod", "D": "The kubelet on node-b needs a restart to discover the volume"}, "answer": "A", "explanation": "never attach it. Fix: use ReadWriteMany (with a supporting provisioner) or run the workload as a StatefulSet with volumeClaimTemplates giving each pod its own volume EXAM TRAP: access modes are about which nodes can attach the volume, not how many pods. ReadWriteOnce means a single node may mount it read-write — the second node's attach will fail forever, which is exactly the stuck ContainerCreating you see. The fix depends on intent: shared storage across nodes needs ReadWriteMany (and a provisioner that supports it, like NFS or EFS), while per-pod storage for replicated workloads is the StatefulSet plus volumeClaimTemplates pattern. Bumping PVC size or restarting kubelets cannot change attach semantics.", "domain": "Domain 4"}
{"id": "bb-05021", "exam": "Subnetting Mastery", "exam_code": "SUBNET", "product": "Study Guide", "question": "For 192.168.200.0/23, what is the broadcast address and how many usable hosts?", "options": {"A": "192.168.201.255 with 510 usable hosts", "B": "192.168.200.255 with 510 usable hosts", "C": "192.168.201.255 with 512 usable hosts", "D": "192.168.200.255 with 254 usable hosts"}, "answer": "A", "explanation": "/23 block 2 in the third octet; 200 is a multiple range .200.0–.201.255; broadcast .201.255; 2^9 - 2 → → = 510 usable. ByteBar · Subnetting Mastery Study Guide — for personal use only 31 BYTEBAR · SWEET IT SKILLS, NO FLUFF", "domain": null}
{"id": "bb-03187", "exam": "Docker Certified Associate", "exam_code": "DCA", "product": "Practice Pack", "question": "A CI pipeline needs to pull a private base image during docker build, using a registry password. The engineer proposes ARG REGISTRY_PASSWORD in the Dockerfile. Why is this dangerous, and what's the safe alternative?", "options": {"A": "There is no risk; ARG is the documented way to pass secrets", "B": "ARG values are printed in build logs; use ENV instead", "C": "ARG values persist in the image's history/metadata (docker history shows them) — secrets leak into every layer and anyone with the image can read them. Use BuildKit secret mounts (--mount=type=secret) so the secret never lands in a layer", "D": "ARG is only for strings under 10 characters; passwords need ENV"}, "answer": "C", "explanation": "every layer and anyone with the image can read them. Use BuildKit secret mounts (--mount=type=secret) so the secret never lands in a layer ARG is build-time only (not persisted into running containers) — but it IS baked into the image metadata: docker history and docker inspect reveal ARG values to anyone holding the image. Secrets in ARG/ENV = secrets in the image forever. BuildKit's --mount=type=secret exposes the secret as a file during a single RUN step without creating a layer containing it. Same family of trap: never bake secrets into images, never pass via -e (visible in docker inspect) — use proper secret mechanisms.", "domain": "Domain 2"}
{"id": "bb-05412", "exam": "TEAS 7", "exam_code": "TEAS-7", "product": "Study Guide", "question": "Choose the most concise revision of \"Due to the fact that she was ill, she missed the exam.\"", "options": {"A": "Because she was ill, she missed the exam.", "B": "Due to the fact of her illness, the exam was missed by her.", "C": "She, being ill, missed the exam due to fact.", "D": "The exam was missed because of the fact she was ill."}, "answer": "A", "explanation": "\"Due to the fact that\" is wordy — \"because\" says the same thing in one word.", "domain": null}
{"id": "bb-02197", "exam": "ISC2 CCSP", "exam_code": "CCSP", "product": "Study Guide", "question": "Which NIST SP 500-292 actor provides the network connectivity and transport that carry cloud services to consumers?", "options": {"A": "Cloud carrier", "B": "Cloud broker", "C": "Cloud auditor", "D": "Cloud provider"}, "answer": "A", "explanation": "The cloud carrier (A) supplies the network, telecommunications, and transport that carry cloud services. The broker (B) is wrong because it negotiates and manages service relationships. The auditor (C) is wrong because it independently assesses controls. The provider (D) is wrong because it makes services available but does not necessarily carry the transport layer.", "domain": null}
{"id": "bb-04323", "exam": "CompTIA PenTest+", "exam_code": "PT0-003", "product": "Study Guide", "question": "The client wants ongoing vulnerability scanning but has no budget for commercial tools. Which open-source scanner do you recommend?", "options": {"A": "Aircrack-ng", "B": "Burp Suite", "C": "Metasploit", "D": "OpenVAS"}, "answer": "D", "explanation": "OpenVAS (Greenbone) is the leading open-source vulnerability scanner — the standard answer when a question asks for a free Nessus alternative. Burp Suite is a web proxy, Metasploit an exploitation framework, Aircrack-ng a wireless auditing suite.", "domain": null}
{"id": "bb-04214", "exam": "CompTIA PenTest+", "exam_code": "PT0-003", "product": "Practice Pack", "question": "During a physical penetration test, what should a tester carry at all times?", "options": {"A": "A lockpick set registered with the local police", "B": "The client's employee badge", "C": "A signed authorization letter", "D": "A copy of the vulnerability scan report"}, "answer": "C", "explanation": "The signed authorization letter — often called a get-out-of-jail-free letter — is proof that the physical intrusion attempts are sanctioned. If building security or law enforcement challenges the tester, producing a letter signed by someone with authority to grant access resolves the situation immediately. A lockpick set proves nothing about permission. Using the client's employee badge is impersonation and creates its own legal problems. The vulnerability scan report belongs to a later technical phase and does not authorize physical access.", "domain": null}
{"id": "bb-01738", "exam": "ISC2 Certified in Cybersecurity", "exam_code": "CC", "product": "Study Guide", "question": "A system grants access only when the login occurs between 8:00 AM and 6:00 PM on weekdays. Basing the decision on this time rule is:", "options": {"A": "Role-based access control", "B": "Mandatory access control", "C": "Discretionary access control", "D": "Rule-based access control"}, "answer": "D", "explanation": "Rule-based access applies global rules (time, location, device) regardless of who you are. RBAC keys off job role; MAC uses labels; DAC lets owners decide. ByteBar · ISC2 CC (Certified in Cybersecurity) Study Guide — Sweet IT Skills. No Fluff.", "domain": null}
{"id": "bb-01381", "exam": "Microsoft AZ-802", "exam_code": "AZ-802", "product": "Study Guide", "question": "You need to present block storage from a dedicated SAN appliance to a Windows Server file server with redundant paths, so a single cable or switch failure does not drop storage access. Which feature and policy should you configure?", "options": {"A": "Multipath I/O (MPIO) with a load-balance policy such as Round Robin", "B": "NIC Teaming with LACP on the storage network", "C": "DFS Replication between the SAN and the server", "D": "iSCSI Initiator with CHAP only"}, "answer": "A", "explanation": "Multipath I/O is Windows Server's framework for redundant storage paths: you install the MPIO feature, claim the iSCSI (or Fibre Channel) devices, and set a load-balance policy like Round Robin so I/O flows across all healthy paths and automatically fails over if a cable, NIC, or switch dies. NIC Teaming (B) aggregates network links but does not provide storage-path failover or per-LUN path policy — MPIO is the storage-correct layer. DFS Replication (C) copies files between servers; it is not block storage multipathing. The iSCSI Initiator with CHAP (D) handles authentication only, with a single path still being a single point of failure.", "domain": null}
{"id": "bb-02904", "exam": "CompTIA Cloud+", "exam_code": "CV0-004", "product": "Practice Pack", "question": "A company keeps three copies of its backups: one on local disk, one on cloud object storage, and one on tape stored offsite. This follows which backup best practice?", "options": {"A": "The 6 Rs of migration", "B": "Incremental forever", "C": "The 3-2-1 rule", "D": "Immutable infrastructure"}, "answer": "C", "explanation": "The 3-2-1 rule calls for 3 copies of data, on 2 different media types, with 1 copy offsite — exactly what this company does. The 6 Rs are migration strategies, incremental-forever is a backup method, and immutable infrastructure is about server builds.", "domain": "Domain 3"}
{"id": "bb-00132", "exam": "Google Professional ML Engineer", "exam_code": null, "product": "Study Guide", "question": "A model's business KPIs slip gradually between quarterly retrainings, but no alert fires. What evaluation practice would catch this?", "options": {"A": "One-time evaluation before deployment", "B": "Evaluating only when retraining", "C": "Trusting training metrics indefinitely", "D": "Continuous evaluation: regularly score the model on fresh labeled data"}, "answer": "D", "explanation": "Continuous evaluation scores the deployed model on fresh labeled data on a schedule, catching decay between retrainings. One-time eval (A) goes stale. Eval-only-at-retrain (B) misses decay windows. Training metrics (C) never reflect production. The MLOps loop: monitor → evaluate continuously → retrain when needed.", "domain": null}
{"id": "bb-05013", "exam": "Subnetting Mastery", "exam_code": "SUBNET", "product": "Study Guide", "question": "Continuing: 10.70.0.0/23 and 10.70.2.0/25 are taken. Where does the 60-host LAN go?", "options": {"A": "10.70.2.64/26", "B": "10.70.2.128/26", "C": "10.70.3.0/26", "D": "10.70.2.128/27"}, "answer": "B", "explanation": "The /25 consumed .2.0–.2.127 → next free 10.70.2.128; 60 → 2^6 - 2 = 62 >= 60 → /26. A is inside the /25; D is too small.", "domain": null}
{"id": "bb-05482", "exam": "CompTIA Linux+", "exam_code": "XK0-006", "product": "Practice Pack", "question": "A company hosts a web server at private IP 10.0.0.5 and must make it reachable from the internet through the firewall's public IP. Which NAT type rewrites the destination address to accomplish this?", "options": {"A": "DNAT (Destination NAT / port forwarding)", "B": "SNAT (Source NAT)", "C": "MASQUERADE", "D": "PAT (Port Address Translation) for outbound sessions"}, "answer": "A", "explanation": "DNAT rewrites the destination address of inbound packets — the mechanism behind port forwarding to internal servers. SNAT/MASQUERADE rewrite the source address of outbound traffic so private hosts can reach the internet. Direction is the key: inbound to a server = DNAT, outbound from clients = SNAT.", "domain": "Domain 3"}
{"id": "bb-03397", "exam": "Family Nurse Practitioner", "exam_code": "FNP", "product": "Study Guide", "question": "A 24-year-old woman requests contraception. She has migraine with aura and smokes. The most appropriate counseling is:", "options": {"A": "Combined oral contraceptives are safe with monitoring", "B": "Barrier methods are the only option", "C": "Depot medroxyprogesterone is contraindicated", "D": "Avoid estrogen-containing methods; consider IUD or implant"}, "answer": "D", "explanation": "Migraine with aura and smoking are contraindications to estrogen due to stroke/VTE risk. Long-acting reversible contraception (IUD, implant) is first-line per guidelines.", "domain": null}
{"id": "bb-01557", "exam": "Microsoft Azure Fundamentals", "exam_code": "AZ-900", "product": "Study Guide", "question": "An employee opens a browser, logs in, and starts using email and spreadsheets with nothing installed on the laptop. Which service type is this?", "options": {"A": "IaaS", "B": "SaaS", "C": "PaaS", "D": "Serverless"}, "answer": "B", "explanation": "Finished software accessed via a browser with nothing to install is SaaS , such as Microsoft 365. Serverless is a compute model where you deploy your own code, not finished applications.", "domain": "Domain 1"}
{"id": "bb-01373", "exam": "Microsoft AZ-802", "exam_code": "AZ-802", "product": "Study Guide", "question": "You need offsite backup of file shares and system state from an on-premises file server to Azure, with the ability to restore individual files from 30 days ago. Which agent and vault should you use?", "options": {"A": "The Azure Backup MARS agent, because it backs up files, folders, and system state directly to a Recovery Services vault", "B": "Azure Site Recovery, because it is designed for file-level backup", "C": "Windows Server Backup to tape only", "D": "DFS Replication to Azure Files"}, "answer": "A", "explanation": "The MARS (Microsoft Azure Recovery Services) agent is the file-and-folder backup tool: installed on the server, it backs up files, folders, and system state on a schedule (up to three times daily) to a Recovery Services vault with configurable retention — and restores are granular down to individual files. ASR (B) replicates whole VMs for disaster recovery; it is not a file-level backup product. Windows Server Backup (C) is local-only unless paired with offsite copies. DFS Replication (D) is not a backup at all — it has no retention, no point-in-time recovery, and deletions replicate too.", "domain": null}
{"id": "bb-02480", "exam": "Certified Kubernetes Application Developer", "exam_code": "CKAD", "product": "Practice Pack", "question": "You need a pod where the app reads LOG_LEVEL from a ConfigMap as an environment variable, reads app.properties from the same ConfigMap as a file at /etc/config, and overrides one key via a command-line flag. Which combination of ConfigMap consumption modes does this require?", "options": {"A": "envFrom for everything — it covers files and flags too", "B": "A volume mount for everything — environment variables can only come from Secrets", "C": "env for the single variable, a volume mount for the file, and args for the flag — the three modes compose in one pod spec", "D": "Three separate pods, one per consumption mode"}, "answer": "C", "explanation": "in one pod spec The three consumption modes are designed to be mixed: env (single key via configMapKeyRef), volumes (keys become files), and plain args for flags. envFrom loads all keys as env vars but cannot produce files. ConfigMaps feed env vars just as well as Secrets do — Secrets are only required when the data is sensitive.", "domain": "Domain 1"}
{"id": "bb-00482", "exam": "Cisco CCNA", "exam_code": "200-301", "product": "Study Guide", "question": "Which command shows the administrative distance and metric of installed routes?", "options": {"A": "show ip route", "B": "show interfaces", "C": "show vlan brief", "D": "show cdp neighbors"}, "answer": "A", "explanation": "show ip route displays each route's [AD/metric], e.g., [110/2] for OSPF.", "domain": null}
{"id": "bb-05523", "exam": "CompTIA Linux+", "exam_code": "XK0-006", "product": "Practice Pack", "question": "'ethtool -S eth0' shows steadily rising rx_dropped counters and you suspect a duplex mismatch. Which command verifies the link's speed and duplex settings?", "options": {"A": "ethtool eth0", "B": "ping -f eth0", "C": "ip route show", "D": "ss -tuln ByteBar · XK0-006 Practice Pack — for personal use only 23"}, "answer": "A", "explanation": "ethtool eth0 reports negotiated speed, duplex, and link status — a half/full mismatch causes collisions, drops, and terrible throughput. Fix by setting both ends identically (prefer auto-negotiation). ping, routing tables, and socket listings cannot show link-layer negotiation. ByteBar · XK0-006 Practice Pack — for personal use only 38 Score yourself Score What it means Next step 85–100 Exam-ready Book the exam. Light review of misses only. 75–84 Close Re-study your two weakest domains, then retake this pack. 60–74 Gaps Rebuild fundamentals in the domains you missed most, then retake. Below 60 Not yet Study core concepts first and give it two more weeks before retaking. Domain answer map Questions 1–23: Domain 1: System Management · Questions 24–43: Domain 2: Services and User Management · Questions 44–61: Domain 3: Security · Questions 62–78: Domain 4: Automation, Orchestration, and Scripting · Questions 79–100: Domain 5: Troubleshooting EXAM TIP — Your weakest domain by percentage — not by raw misses — is what to study next. 85+ and you're ready. Go earn it. ByteBar · XK0-006 Practice Pack — for personal use only 39 BYTEBAR · SWEET IT SKILLS, NO FLUFF A final note ByteBar is an independent study resource and is not affiliated with or endorsed by CompTIA, Inc. All 100 questions in this pack are original practice questions written by ByteBar for study purposes. They are not real exam questions, recalled exam content, or exam dumps, and they do not guarantee that you will pass the XK0-006 exam. CompTIA, Linux+, and related marks are trademarks of CompTIA, Inc. This product is an independent study aid — it is not sponsored, endorsed, or approved by CompTIA. Good luck — and go earn it. ByteBar · XK0-006 Practice Pack — for personal use only 40", "domain": "Domain 5"}
{"id": "bb-05322", "exam": "TEAS 7", "exam_code": "TEAS-7", "product": "Practice Pack", "question": "Gas exchange between air and blood occurs in the lungs at the:", "options": {"A": "Bronchi, which branch into the lungs.", "B": "Alveoli, whose thin walls sit against dense capillary networks.", "C": "Trachea, which carries air from the throat.", "D": "Diaphragm, which powers breathing movements."}, "answer": "B", "explanation": "B is correct because alveoli are microscopic air sacs with walls one cell thick, wrapped in capillaries — oxygen diffuses from alveolar air into blood while carbon dioxide diffuses out, driven by concentration gradients across that minimal barrier. A is wrong because bronchi are conducting airways; they transport air but their walls are too thick for significant gas exchange. C is wrong because the trachea is a rigid windpipe for airflow, not an exchange surface. D is wrong because the diaphragm is a muscle that drives ventilation; it moves air but performs no gas exchange itself.", "domain": null}
{"id": "bb-04007", "exam": "PMHNP-BC", "exam_code": "PMHNP-BC", "product": "Practice Pack", "question": "A patient asks how disulfiram helps maintain alcohol abstinence. Which mechanism is correct?", "options": {"A": "It reduces alcohol craving by blocking mu-opioid receptors", "B": "It restores glutamate-GABA balance during early abstinence", "C": "It produces mild opioid agonist effects that blunt withdrawal", "D": "It blocks aldehyde dehydrogenase, causing an aversive reaction if alcohol is consumed"}, "answer": "D", "explanation": "Disulfiram irreversibly inhibits aldehyde dehydrogenase, the enzyme that converts acetaldehyde to acetate during alcohol metabolism. If the patient drinks, acetaldehyde accumulates and triggers flushing, throbbing headache, nausea, vomiting, hypotension, and tachycardia within minutes, which creates the intended deterrent. It does not reduce craving the way naltrexone does, it does not enhance GABA like acamprosate, and it has no opioid activity whatsoever. Because the reaction can be severe, patients must be abstinent before starting, understand that even small amounts of alcohol in food or products can trigger it, and give informed consent, making patient selection and education central to safe prescribing.", "domain": null}
{"id": "bb-00449", "exam": "Cisco CCNA", "exam_code": "200-301", "product": "Study Guide", "question": "On a router-on-a-stick, which subinterface command tags traffic for VLAN 20?", "options": {"A": "encapsulation dot1Q 20", "B": "switchport access vlan 20", "C": "vlan 20 encapsulation isl", "D": "dot1q trunk native 20"}, "answer": "A", "explanation": "'encapsulation dot1Q 20' on the subinterface (e.g., Gi0/0.20) tags its traffic for VLAN 20. Switchport commands belong on switches, and ISL is Cisco's legacy trunking protocol.", "domain": "Switching & Network Access"}
{"id": "bb-01418", "exam": "Microsoft Azure Fundamentals", "exam_code": "AZ-900", "product": "Practice Pack", "question": "A mobile game needs a database that serves players worldwide with single-digit-millisecond reads and automatic global distribution. Which Azure service fits BEST?", "options": {"A": "Azure Files", "B": "Azure Blob storage", "C": "Azure Cosmos DB", "D": "Azure Data Box"}, "answer": "C", "explanation": "Cosmos DB is the globally distributed, multi-model database with turnkey worldwide replication and millisecond latency. Blob/Files are storage, not databases; Data Box is offline transfer.", "domain": "Domain 2"}
{"id": "bb-00942", "exam": "AWS Certified AI Practitioner", "exam_code": "AIF-C01", "product": "Practice Pack", "question": "For high-stakes AI decisions (such as loan approvals), which best practice is most important?", "options": {"A": "Hiding the AI's involvement from applicants", "B": "Using the largest model regardless of evaluation", "C": "Fully automating with no human review to save time", "D": "Keeping a human in the loop to review and oversee AI-influenced decisions"}, "answer": "D", "explanation": "Human oversight - human-in-the-loop review - is critical for high-stakes decisions. It catches model errors and bias, ensures accountability, and supports regulatory compliance.", "domain": "Domain 4"}
{"id": "bb-02449", "exam": "Certified Kubernetes Application Developer", "exam_code": "CKAD", "product": "Practice Pack", "question": "The task says: 'Don't send traffic to the pod until it's warm, but restart it if it deadlocks later.' Which two probes implement this, and what does each control?", "options": {"A": "Two livenessProbes — one with a long delay for warmup, one aggressive for deadlocks", "B": "readinessProbe for 'don't send traffic until warm' (it gates Service endpoints) and livenessProbe for 'restart if deadlocked' (it triggers container restarts) — they serve different masters", "C": "A startupProbe for both — it handles warmup and deadlocks", "D": "readinessProbe for both — failing readiness restarts the container"}, "answer": "B", "explanation": "'restart if deadlocked' (it triggers container restarts) — they serve different masters The fundamental split: readiness gates TRAFFIC (failing readiness removes the pod from endpoints, no restart), liveness gates RESTARTS (failing liveness kills and restarts the container). 'Don't send traffic until warm' = readiness; 'restart if wedged' = liveness. Failing readiness never restarts anything — confusing the two is the exam's core probe trap.", "domain": "Domain 5"}
{"id": "bb-01788", "exam": "ISC2 Certified in Cybersecurity", "exam_code": "CC", "product": "Study Guide", "question": "When the ideal security control is not feasible, the alternative control put in its place is called a:", "options": {"A": "Directive control", "B": "Recovery control", "C": "Deterrent control", "D": "Compensating control"}, "answer": "D", "explanation": "A compensating control is an acceptable substitute when the primary control can't be implemented — classic exam scenario: a legacy system that can't be encrypted gets extra monitoring instead.", "domain": null}
{"id": "bb-04617", "exam": "AWS Solutions Architect Associate", "exam_code": "SAA-C03", "product": "Practice Pack", "question": "A company wants continuous, ML-powered threat detection across CloudTrail, VPC Flow Logs, and DNS logs without deploying agents. Which service provides this?", "options": {"A": "AWS WAF", "B": "Amazon Inspector", "C": "Amazon GuardDuty", "D": "AWS Artifact"}, "answer": "C", "explanation": "GuardDuty analyzes CloudTrail, Flow Logs, and DNS logs with machine learning to detect threats, with no agents required. WAF filters web requests, Inspector scans for software vulnerabilities, and Artifact provides compliance reports. ByteBar · SAA-C03 Practice Pack — for personal use only 34", "domain": "Domain 3"}
{"id": "bb-00490", "exam": "Cisco CCNA", "exam_code": "200-301", "product": "Study Guide", "question": "Which port-security violation mode drops offending traffic AND sends an SNMP trap?", "options": {"A": "Protect", "B": "Restrict", "C": "Shutdown ByteBar · Cisco CCNA (200-301) Study Guide — Sweet IT Skills. No Fluff.", "D": "Disabled"}, "answer": "B", "explanation": "Restrict drops + logs/traps but keeps the port up. Protect only drops; shutdown (default) err-disables the port.", "domain": null}
{"id": "bb-00307", "exam": "Cisco CCNA", "exam_code": "200-301", "product": "Practice Pack", "question": "A network 10.10.0.0/16 must be split into subnets with 500 hosts each. Which mask fits best?", "options": {"A": "/23 (510 hosts)", "B": "/24 (254 hosts)", "C": "/22 (1022 hosts)", "D": "/25 (126 hosts)"}, "answer": "A", "explanation": "/23 gives 510 usable hosts — the smallest mask meeting the 500-host requirement with minimal waste.", "domain": "D1"}
{"id": "bb-04471", "exam": "Red Hat RHCSA/RHCE", "exam_code": "RHCSA", "product": "Practice Pack", "question": "Practice scenario: you SSH daily as 'admin' to db01.example.com using ~/.ssh/db_key. Which client-side config removes the need for '-i' every time?", "options": {"A": "Add alias sshdb='ssh -i ~/.ssh/db_key' to ~/.bashrc", "B": "In ~/.ssh/config add a block: Host db01 / HostName db01.example.com / User admin / IdentityFile ~/.ssh/db_key", "C": "Append the key path to ~/.ssh/known_hosts", "D": "Append \"IdentityFile ~/.ssh/db_key\" to /etc/ssh/ssh_config ByteBar · RHCSA / RHCE Practice Pack · for personal use only 14 Ansible Automation — 20 questions"}, "answer": "B", "explanation": "~/.ssh/db_key B is correct: ~/.ssh/config Host blocks are the standard per-user mechanism - HostName, User, and IdentityFile travel together and apply automatically to matching hosts. A edits the system-wide config, affecting all users and requiring root - the wrong scope for a personal key. B works but is fragile shell trickery that breaks scp, rsync, and Ansible invocations. D is nonsense: known_hosts stores server host keys, never client identity paths. ByteBar · RHCSA / RHCE Practice Pack · for personal use only 34", "domain": "Networking, Users & Security"}
{"id": "bb-03749", "exam": "CompTIA Network+", "exam_code": "N10-009", "product": "Study Guide", "question": "95 — A switch log shows the same access port cycling up and down every few seconds for an hour. What is this symptom called, and what is the FIRST physical check?", "options": {"A": "Port flapping - check the cable and NIC for a bad connection", "B": "A routing loop - check the routing table", "C": "A broadcast storm - check for STP", "D": "MAC flooding - check the CAM table"}, "answer": "A", "explanation": "Rapid up/down cycling is port flapping - usually a marginal cable, bad NIC, or duplex/speed mismatch. Swap the cable and test the NIC before touching the config.", "domain": null}
{"id": "bb-03131", "exam": "CompTIA SecAI+", "exam_code": "CY0-001", "product": "Study Guide", "question": "A team wants to regression-test their support chatbot's prompts every time the underlying model is updated, comparing answer quality and safety across versions. Which tool is built for this?", "options": {"A": "Promptfoo, because it runs repeatable evaluations of prompts across models and versions", "B": "Garak, because it scans networks for misconfigurations", "C": "Metasploit, because it exploits operating-system vulnerabilities", "D": "Wireshark, because it captures the chatbot's network traffic"}, "answer": "A", "explanation": "Promptfoo is correct because it is designed for systematic prompt evaluation: running the same test cases across prompt versions and model updates to catch regressions in quality or safety. Garak is wrong because it probes for vulnerabilities rather than measuring prompt quality over time. Metasploit is wrong because it is an exploitation framework for conventional systems, not a prompt-testing harness. Wireshark is wrong because packet capture cannot evaluate whether prompt responses are correct or safe.", "domain": null}
{"id": "bb-04929", "exam": "Subnetting Mastery", "exam_code": "SUBNET", "product": "Practice Pack", "question": "What decimal value does the binary number 11100000 represent?", "options": {"A": "224", "B": "192", "C": "240", "D": "208"}, "answer": "A", "explanation": "128 + 64 + 32 = 224. Three 1s from the left is always 224.", "domain": null}
{"id": "bb-02454", "exam": "Certified Kubernetes Application Developer", "exam_code": "CKAD", "product": "Practice Pack", "question": "The task says: 'Install the redis chart from Bitnami with 3 replicas.' You run helm repo add bitnami then helm install myredis bitnami/redis --set replica.replicaCount=3, and get 'chart \"redis\" not found'. What step did you skip?", "options": {"A": "helm dependency update on the redis chart", "B": "helm pull — charts must be downloaded first", "C": "helm search — you must search before installing", "D": "helm repo update — the local index is stale after adding a repo; update refreshes it before install ByteBar · CKAD Practice Pack — for personal use only 15"}, "answer": "D", "explanation": "helm repo add only registers the repo; the local chart index is empty/stale until helm repo update downloads it. Skipping update is the #1 Helm exam trap — 'chart not found' right after adding the repo. The winning sequence is always: repo add → repo update → show values → install.", "domain": "Domain 3"}
{"id": "bb-03979", "exam": "NCLEX-RN", "exam_code": "NCLEX-RN", "product": "Study Guide", "question": "Which newly prescribed medication most increases an older adult’s fall risk?", "options": {"A": "A daily multivitamin", "B": "A sedative-hypnotic for sleep", "C": "A topical antifungal", "D": "Artificial tears"}, "answer": "B", "explanation": "Sedative-hypnotics top the fall-risk medication list — sedation, slowed reflexes, orthostatics, and nighttime bathroom trips are the perfect storm. Multivitamins (A), topicals (C), and artificial tears (D) don’t touch balance or alertness. Review the whole med list for culprits: antihypertensives, diuretics, opioids, and anything anticholinergic.", "domain": null}
{"id": "bb-01720", "exam": "ISC2 Certified in Cybersecurity", "exam_code": "CC", "product": "Study Guide", "question": "A firewall that tracks active connections and automatically allows legitimate return traffic is a:", "options": {"A": "Stateful firewall", "B": "Stateless firewall", "C": "Host-based firewall", "D": "Proxy firewall"}, "answer": "A", "explanation": "Stateful inspection remembers connection state — return packets for an established session pass without explicit rules. Stateless filters judge every packet in isolation and need mirrored rules for replies. ByteBar · ISC2 CC (Certified in Cybersecurity) Study Guide — Sweet IT Skills. No Fluff.", "domain": null}
{"id": "bb-01269", "exam": "Microsoft AZ-802", "exam_code": "AZ-802", "product": "Practice Pack", "question": "The help-desk team needs to unlock user accounts that get locked out, but for compliance reasons they must NOT be able to reset passwords or change any other account attributes. The accounts live in the CorpUsers OU. How should you delegate this?", "options": {"A": "Create a custom delegation on the CorpUsers OU granting the help-desk group Write access to the lockoutTime property only", "B": "Add the help-desk group to the Account Operators built-in group", "C": "Use the Delegation of Control Wizard to grant Reset user passwords and force password change at next logon", "D": "Add the help-desk group to the Domain Admins group with a documented exception"}, "answer": "A", "explanation": "Create a custom delegation on the CorpUsers OU granting the help-desk group Write access to the lockoutTime property only — Unlocking an account is a write to the lockoutTime attribute, so a custom (property-specific) delegation granting only that write gives the help desk exactly the unlock ability and nothing else: least privilege. Account Operators (B) can manage most accounts domain-wide, far beyond unlocking. The wizard's password-reset task (C) grants the very capability compliance forbids. Domain Admins (D) is the opposite of least privilege. This is the exam pattern for surgical delegation: when no wizard task matches, delegate the specific property.", "domain": null}
{"id": "bb-02921", "exam": "CompTIA Cloud+", "exam_code": "CV0-004", "product": "Practice Pack", "question": "A deployment pipeline fails at the deploy stage with no clear error in the console. Where should the administrator look FIRST for details?", "options": {"A": "The company's social media feed", "B": "The detailed deployment and application logs", "C": "The physical security camera footage", "D": "The employee handbook"}, "answer": "B", "explanation": "Deployment and application logs contain the detailed error output that consoles summarize or omit — the first place to diagnose a pipeline failure. Social media, camera footage, and handbooks contain no deployment diagnostics.", "domain": "Domain 6"}
{"id": "bb-01088", "exam": "AWS Certified AI Practitioner", "exam_code": "AIF-C01", "product": "Study Guide", "question": "The key innovation of the transformer architecture is:", "options": {"A": "The self-attention mechanism", "B": "Backpropagation", "C": "Decision forests", "D": "Token billing"}, "answer": "A", "explanation": "Transformers - the architecture behind modern LLMs - use self-attention to weigh the importance of different words in context.", "domain": "Domain 2"}
{"id": "bb-04801", "exam": "Microsoft SC-500", "exam_code": "SC-500", "product": "Study Guide", "question": "Contoso has legacy servers that cannot be patched, and the board formally accepts the risk for 12 months. How should the security team reflect this decision so the servers stop dragging down the secure score and compliance results?", "options": {"A": "Delete the unhealthy recommendations", "B": "Suppress each related alert in Defender XDR", "C": "Disable the Defender for Servers plan on the subscription", "D": "Create a time-bound exemption scoped to the servers with the risk-accepted category"}, "answer": "D", "explanation": "Exemptions remove a scope, such as a management group, subscription, resource group, or resource, from secure score calculations and compliance assessments, with a risk-accepted or mitigated category and an expiry date forcing re-review. A 12-month board-accepted risk is the textbook risk-accepted use case. Deleted recommendations regenerate. Alert suppression in Defender XDR affects incident noise, not posture scoring. Disabling the Defender plan removes protection and visibility instead of documenting an accepted exception.", "domain": null}
{"id": "bb-02767", "exam": "AWS Certified Cloud Practitioner", "exam_code": "CLF-C02", "product": "Study Guide", "question": "An e-commerce site automatically adds web servers during a flash sale and removes them afterward. Which concept does this describe?", "options": {"A": "Elasticity", "B": "Scalability", "C": "Agility", "D": "Fault tolerance"}, "answer": "A", "explanation": "Elasticity = resources automatically expand and contract with demand. Scalability is the ability to handle growth; agility is speed of innovation; fault tolerance is surviving component failure.", "domain": null}
{"id": "bb-02083", "exam": "ISC2 CCSP", "exam_code": "CCSP", "product": "Practice Pack", "question": "Cloud Security Operations An attacker has a foothold on several cloud VMs and is moving laterally. Which containment approach is BEST?", "options": {"A": "Delete all affected systems immediately", "B": "Isolate affected systems short-term, then rebuild from known-good images long-term", "C": "Ignore it and monitor", "D": "Pay any ransom demanded"}, "answer": "B", "explanation": "Sound containment (B) pairs short-term isolation (stop the bleeding: disconnect, block) with long-term containment (rebuild from known-good media so the attacker has no foothold). Deleting systems (A) destroys forensic evidence. Ignoring (C) lets the attacker persist. Paying ransom (D) funds crime and does not guarantee recovery.", "domain": "Cloud Security Operations"}
{"id": "bb-03054", "exam": "CompTIA SecAI+", "exam_code": "CY0-001", "product": "Practice Pack", "question": "An analyst extracts readable strings from an unpacked stealer binary and asks AI to pull out candidate C2 domains and IPs. What must happen before any are blocked?", "options": {"A": "Block every string the AI returns, because speed beats accuracy", "B": "Trust all AI-extracted IOCs, because AI does not make mistakes", "C": "Validate each candidate IOC (reputation, context, age) before blocking, because AI extraction can surface benign or stale indicators", "D": "Skip validation, because the binary is definitely malicious"}, "answer": "C", "explanation": "Validating each candidate IOC is correct because AI string extraction is noisy: it surfaces hardcoded test domains, shared-library artifacts, and long-dead infrastructure alongside real C2, and blocking benign indicators causes outages and false alarms. Blocking everything is wrong because speed without accuracy turns defense into self-inflicted denial of service. Trusting all AI output is wrong because extraction models hallucinate and misclassify; they are triage aids, not oracles. Skipping validation because the binary is malicious is wrong because a malicious binary still contains benign strings — malice of the source does not purify the output.", "domain": null}
{"id": "bb-01222", "exam": "Microsoft AZ-802", "exam_code": "AZ-802", "product": "Practice Pack", "question": "Compliance requires monthly full backups of file servers kept for 7 years. Keeping seven years of recovery points in the default vault tier would be far too expensive. Which Recovery Services vault capability provides low-cost long-term retention?", "options": {"A": "The operational tier — it keeps snapshots for 7 years at no extra cost", "B": "Zone-redundant storage — it discounts long retention", "C": "Soft delete — it extends retention to 7 years automatically", "D": "The vault-archive tier — move long-term recovery points to archive storage, which is much cheaper with a retrieval latency of several hours"}, "answer": "D", "explanation": "The vault-archive tier — move long-term recovery points to archive storage, which is much cheaper with a retrieval latency of several hours — Vault-archive is the cold tier for Recovery Services vaults: you move older recovery points there for drastically lower storage cost, accepting retrieval times measured in hours — the right trade-off for 7-year compliance retention you will rarely touch. The operational tier (A) is short-term snapshot storage for instant restores, not multi-year retention. Soft delete (C) is a 14-day safety net against accidental deletion, not a retention tier. Redundancy options (D) affect durability and availability, not retention pricing.", "domain": null}
{"id": "bb-03101", "exam": "CompTIA SecAI+", "exam_code": "CY0-001", "product": "Study Guide", "question": "A developer strips control characters and instruction-like phrases such as ignore previous instructions from user input before it is sent to the company's LLM. Which defensive control is this?", "options": {"A": "Input validation and sanitization, because untrusted user input must be checked before it reaches the model", "B": "Output filtering, because the response is where the damage appears", "C": "Rate limiting, because too many requests cause injection", "D": "Model encryption, because encrypted prompts cannot be abused"}, "answer": "A", "explanation": "Input validation and sanitization is correct because prompt injection arrives through user input, so stripping control characters, instruction-like phrases, and unexpected formats at the boundary is the direct defense. Output filtering is wrong because it inspects what the model produces, not what the attacker feeds in. Rate limiting is wrong because it slows abuse but does not stop a single malicious prompt from working. Model encryption is wrong because encryption protects data at rest or in transit; it does nothing to neutralize malicious instructions inside a prompt.", "domain": null}
{"id": "bb-04582", "exam": "Red Hat RHCSA/RHCE", "exam_code": "RHCSA", "product": "Study Guide", "question": "You want to use a systemd timer (not cron) to run a backup every night. Which pair of unit files do you need to create?", "options": {"A": "mybackup.service and mybackup.cron", "B": "mybackup.timer and mybackup.target", "C": "mybackup.service and mybackup.timer", "D": "An entry in /etc/crontab only"}, "answer": "C", "explanation": "C is correct: a systemd timer setup needs a .service unit describing what to run and a .timer unit with the same base name describing when to run it (via OnCalendar=). You then enable and start the timer, not the service: 'systemctl enable --now mybackup.timer'. A is wrong because there is no .cron unit type in systemd. D is wrong because /etc/crontab is the legacy cron mechanism, not a systemd timer. B is wrong because .target units group other units for boot milestones — a timer must be paired with the .service it triggers. Verify timers with 'systemctl list-timers'.", "domain": null}
{"id": "bb-05001", "exam": "Subnetting Mastery", "exam_code": "SUBNET", "product": "Study Guide", "question": "A subnet mask octet reads 11100100 in binary. Its decimal value is:", "options": {"A": "228", "B": "244", "C": "220", "D": "236"}, "answer": "A", "explanation": "128 + 64 + 32 + 4 = 228. B is 11110100, C is 11011100, D is 11101100. ByteBar · Subnetting Mastery Study Guide — for personal use only 21 BYTEBAR · SWEET IT SKILLS, NO FLUFF", "domain": null}
{"id": "bb-01898", "exam": "Adult CCRN", "exam_code": "CCRN", "product": "Practice Pack", "question": "Which patient is NOT an appropriate candidate for bilevel positive airway pressure (BiPAP)?", "options": {"A": "An alert, cooperative patient with COPD exacerbation and hypercapnia", "B": "An alert patient with cardiogenic pulmonary edema", "C": "A patient with active vomiting, inability to protect the airway, and hemodynamic instability", "D": "An alert high-risk patient immediately post-extubation"}, "answer": "C", "explanation": "Noninvasive ventilation demands an alert, cooperative patient who can protect the airway and clear secretions. Contraindications include vomiting or high aspiration risk, facial trauma or burns preventing mask seal, upper airway obstruction, inability to clear secretions, and hemodynamic instability or unstable arrhythmia. Delaying intubation by persisting with BiPAP in such patients increases mortality. COPD exacerbations, cardiogenic pulmonary edema, and post-extubation support in selected alert patients are appropriate indications.", "domain": null}
{"id": "bb-00671", "exam": "CompTIA A+", "exam_code": "220-1201/220-1202", "product": "Study Guide", "question": "What should you do BEFORE editing the Windows Registry?", "options": {"A": "Disable System Restore", "B": "Back up the registry / create a restore point", "C": "Boot into Safe Mode", "D": "Run chkdsk"}, "answer": "B", "explanation": "Always back up first — export the key or create a restore point. A bad registry edit can prevent Windows from booting.", "domain": null}
{"id": "bb-03050", "exam": "CompTIA SecAI+", "exam_code": "CY0-001", "product": "Practice Pack", "question": "An AI service monitors paste sites, forums, and marketplaces for the company's domain credentials and customer data, alerting the SOC to fresh exposures. What is the analyst's required step when an alert fires?", "options": {"A": "Reset every password company-wide, because the AI might be right", "B": "Ignore the alert, because external monitoring is unreliable", "C": "Publish the findings, because transparency deters criminals", "D": "Validate the exposure (is it real, current, ours?) before triggering credential resets, because AI monitoring produces false positives"}, "answer": "D", "explanation": "Validating the exposure first is correct because AI external monitoring is noisy — it flags test credentials, other companies' leaks, and stale dumps; confirming the data is real, current, and actually yours before forcing resets avoids alert fatigue and unnecessary disruption. Resetting every password company-wide is wrong because it is wildly disproportionate to an unvalidated alert and trains users to resent security. Ignoring the alert is wrong because external exposure monitoring is a genuine early-warning source; the answer is validation, not dismissal. Publishing the findings is wrong because publicizing unvalidated breach claims creates legal exposure and panic. Section 4 — AI Governance, Risk & Compliance", "domain": null}
{"id": "bb-04010", "exam": "PMHNP-BC", "exam_code": "PMHNP-BC", "product": "Practice Pack", "question": "A patient has maintained alcohol abstinence for two weeks after detox and wants medication to help stay abstinent. Which statement about acamprosate is correct?", "options": {"A": "It creates an aversive reaction if alcohol is consumed, like disulfiram.", "B": "It treats acute alcohol withdrawal symptoms during detoxification.", "C": "It helps maintain abstinence by modulating glutamate and GABA balance, is dosed three times daily, and requires renal dose adjustment.", "D": "It is given as a monthly injection and works through opioid blockade."}, "answer": "C", "explanation": "Acamprosate is indicated for maintaining abstinence in alcohol-dependent patients already detoxified, working by restoring the glutamate-GABA balance disrupted by chronic alcohol exposure and dampening the protracted withdrawal hyperexcitability that drives relapse. Its thrice-daily dosing demands honest adherence counseling, and because it is renally cleared, dose reduction or avoidance is required in renal impairment. It produces no aversive disulfiram-like reaction; that is a different medication with a different contract. It has no role in acute withdrawal, which is managed with benzodiazepines. Monthly injection with opioid blockade describes extended-release naltrexone, not acamprosate, whose mechanism is glutamatergic rather than opioidergic. Section 9 — Psychotherapy & Related Theories (continued)", "domain": null}
{"id": "bb-02763", "exam": "AWS Certified Cloud Practitioner", "exam_code": "CLF-C02", "product": "Study Guide", "question": "An architect deletes idle EBS volumes and rightsizes over-provisioned EC2 instances. Which Well-Architected Framework pillar does this support?", "options": {"A": "Security", "B": "Reliability", "C": "Cost Optimization", "D": "Sustainability"}, "answer": "C", "explanation": "Eliminating waste and paying only for what you need = Cost Optimization. (Sustainability is about reducing environmental impact — related, but the exam maps waste-reduction-for-the-bill to Cost Optimization.)", "domain": null}
{"id": "bb-00511", "exam": "CompTIA A+", "exam_code": "220-1201/220-1202", "product": "Practice Pack", "question": "A laser printer outputs garbled characters on every job. What is the MOST likely cause?", "options": {"A": "Empty paper tray", "B": "Wrong or corrupt printer driver", "C": "Low toner", "D": "Network congestion"}, "answer": "B", "explanation": "Garbled output = the printer is misinterpreting the data stream — almost always a driver issue.", "domain": null}
{"id": "bb-03493", "exam": "ITIL 4 Foundation", "exam_code": "ITIL-4", "product": "Practice Pack", "question": "A monitoring tool detects that disk usage has reached 85% and sends an alert so action can be taken before the disk fills. This is what type of event?", "options": {"A": "Informational event", "B": "Warning event", "C": "Exception event", "D": "Planned event"}, "answer": "B", "explanation": "A warning event signals that a threshold is approaching and action is needed to prevent a problem. Informational events need no action; exception events indicate abnormal operation, such as a breached threshold.", "domain": "Practices in detail (7)"}
{"id": "bb-05070", "exam": "Subnetting Mastery", "exam_code": "SUBNET", "product": "Study Guide", "question": "Which dotted-decimal mask is exactly a /15?", "options": {"A": "255.252.0.0", "B": "255.255.0.0", "C": "255.254.0.0", "D": "255.248.0.0"}, "answer": "C", "explanation": "/15 = 8 + 7 bits → 11111110 = 254 → 255.254.0.0. A is /14, B is /16, D is /13.", "domain": null}
{"id": "bb-01017", "exam": "AWS Certified AI Practitioner", "exam_code": "AIF-C01", "product": "Study Guide", "question": "A company wants natural-language search across its wikis, SharePoint, and S3 documents with direct answers. Which service fits?", "options": {"A": "Amazon Lex", "B": "Amazon Kendra", "C": "Amazon Personalize", "D": "Amazon Textract"}, "answer": "B", "explanation": "Kendra is intelligent enterprise search: it indexes content across data sources and returns answers - not just document links - to natural-language questions.", "domain": "Domain 3"}
{"id": "bb-05679", "exam": "CompTIA Linux+", "exam_code": "XK0-006", "product": "Study Guide", "question": "You need a GPT partition table on a 4 TB disk. Which tool is purpose-built for GPT?", "options": {"A": "fdisk", "B": "gdisk", "C": "mkfs.ext4", "D": "fsck"}, "answer": "B", "explanation": "gdisk is the GPT partitioning tool (fdisk is the classic MBR-oriented one). mkfs/fsck work on filesystems, not partition tables. ByteBar · CompTIA Linux+ (XK0-006) Study Guide — for personal use only 24", "domain": null}
{"id": "bb-01443", "exam": "Microsoft Azure Fundamentals", "exam_code": "AZ-900", "product": "Practice Pack", "question": "Azure Load Balancer operates at which network layer and what does it do?", "options": {"A": "Layer 2 — switches frames between physical servers", "B": "Layer 3 — assigns public IP addresses to virtual machines", "C": "Layer 7 — inspects HTTP content and applies web application firewall rules", "D": "Layer 4 — distributes TCP/UDP traffic across backend resources for availability and scale"}, "answer": "D", "explanation": "Load Balancer is a Layer 4 (transport) distributor of TCP/UDP flows — high throughput, zone-redundant. Layer 7 smarts (URL routing, WAF) belong to Application Gateway; it doesn't assign IPs or switch frames.", "domain": "Domain 2"}
{"id": "bb-02726", "exam": "AWS Certified Cloud Practitioner", "exam_code": "CLF-C02", "product": "Study Guide", "question": "A company’s AWS bill shows exactly how many EC2 hours and gigabytes of S3 storage it consumed, because usage is metered per service. Which essential characteristic of cloud computing describes this?", "options": {"A": "On-demand self-service", "B": "Broad network access", "C": "Resource pooling BYTEBAR · SWEET IT SKILLS, NO FLUFF", "D": "Measured service"}, "answer": "D", "explanation": "Measured service = resource usage is monitored, controlled, and billed per unit consumed — the basis of pay-as-you-go. The other three characteristics describe how you get and reach resources, not how usage is metered.", "domain": null}
{"id": "bb-01186", "exam": "Microsoft Azure Administrator", "exam_code": "AZ-104", "product": "Study Guide", "question": "You need to host a static marketing website with a custom index document and error page, with no compute resources to manage. What should you configure?", "options": {"A": "An App Service web app on the Free tier", "B": "A CDN endpoint pointing at a file share", "C": "Static website hosting on a general-purpose v2 storage account", "D": "Blob versioning on the $root container"}, "answer": "C", "explanation": "Storage static website hosting serves HTML/CSS/JS directly from the $web container with index and error document support — zero compute. A CDN (B) can front it for performance, but the origin here is the storage static site, not a file share.", "domain": null}
{"id": "bb-03437", "exam": "Family Nurse Practitioner", "exam_code": "FNP", "product": "Study Guide", "question": "A 58-year-old with type 2 diabetes has a blood pressure of 128/78, an eGFR of 58, and an elevated urine albumin-to-creatinine ratio. The best medication addition is:", "options": {"A": "A thiazide diuretic", "B": "An ACE inhibitor", "C": "No medication — the blood pressure is normal", "D": "An NSAID for his knee pain"}, "answer": "B", "explanation": "Albuminuria in diabetes means the kidneys need renoprotection regardless of blood pressure — ACE inhibitors (or ARBs) reduce proteinuria and slow CKD progression even in normotensive patients. A thiazide lowers pressure but does not specifically protect the glomerulus. 'BP is normal so do nothing' ignores the albuminuria indication. NSAIDs in CKD with albuminuria accelerate renal decline — avoid them.", "domain": null}
{"id": "bb-00215", "exam": "HashiCorp Terraform Associate", "exam_code": "004", "product": "Practice Pack", "question": "A configuration declares two `provider \"aws\"` blocks — one for us-east-1 and one aliased `west` for us-west-2. A resource block has no `provider` meta-argument. Which provider configuration will Terraform use for that resource?", "options": {"A": "The apply fails — every resource must declare an explicit provider", "B": "The unaliased (default) provider configuration, us-east-1", "C": "Terraform picks whichever provider block was declared first", "D": "The aliased `west` provider, because aliases take precedence"}, "answer": "B", "explanation": "Resources without an explicit provider meta-argument use the default (unaliased) configuration of that provider type. A is false — the default exists precisely so most resources need no declaration. C and D invent selection rules; the rule is simply default-vs-aliased.", "domain": "Domain 2"}
{"id": "bb-03882", "exam": "NCLEX-RN", "exam_code": "NCLEX-RN", "product": "Practice Pack", "question": "Which action demonstrates correct care of an indwelling urinary catheter?", "options": {"A": "Keeping the drainage bag above the level of the bladder.", "B": "Disconnecting the drainage bag daily to rinse it out.", "C": "Keeping the bag below bladder level with the tubing free of kinks and dependent loops.", "D": "Applying powder around the meatus to reduce moisture."}, "answer": "C", "explanation": "Correct catheter care maintains a closed drainage system with the bag below bladder level so urine flows by gravity, and keeps tubing free of kinks and dependent loops that obstruct flow and invite infection. Option A is wrong because a bag above bladder level allows urine to reflux into the bladder. Option B is wrong because disconnecting the system breaks sterility. Option D is wrong because powder irritates the meatus and promotes bacterial growth.", "domain": null}
{"id": "bb-02845", "exam": "CompTIA CySA+", "exam_code": "CS0-004", "product": "Practice Pack", "question": "What is the MAIN reason to involve legal counsel early in a suspected data breach?", "options": {"A": "To protect sensitive investigation communications under privilege, guide notification obligations, and manage regulatory and litigation risk", "B": "To slow down the investigation", "C": "Legal counsel performs the forensic imaging", "D": "Lawyers enjoy incident response ByteBar · CS0-004 Practice Pack — for personal use only 19"}, "answer": "A", "explanation": "manage regulatory and litigation risk Early legal involvement preserves privilege over investigation findings, ensures notification laws are met, and manages downstream liability. Counsel advises — they do not image drives or deliberately slow things down.", "domain": "Domain 3"}
{"id": "bb-04452", "exam": "Red Hat RHCSA/RHCE", "exam_code": "RHCSA", "product": "Practice Pack", "question": "Practice scenario: install httpd only on Red Hat-family hosts. Which 'when' conditional is correct?", "options": {"A": "when: ansible_facts['os_family'] = \"RedHat\"", "B": "when: ansible_facts['os_family'] == \"RedHat\"", "C": "when: ansible_facts.os_family == \"redhat\"", "D": "when: ansible_os_family == RedHat"}, "answer": "B", "explanation": "B is correct: it uses the canonical facts dict with the exact fact value \"RedHat\" (capital R and H). A fails on case - the fact value is \"RedHat\", not \"redhat\". C leaves RedHat unquoted, so Ansible treats it as a variable name and errors. D uses a single '=', which is assignment, not comparison - a Jinja syntax error. ByteBar · RHCSA / RHCE Practice Pack · for personal use only 35", "domain": "Ansible Automation"}
{"id": "bb-04598", "exam": "AWS Solutions Architect Associate", "exam_code": "SAA-C03", "product": "Practice Pack", "question": "A company needs automated, scheduled EBS snapshots with retention policies and cross-account sharing for DR, without custom scripts. Which service should be used?", "options": {"A": "Amazon S3 lifecycle policies", "B": "AWS Config rules", "C": "Amazon Data Lifecycle Manager (DLM) lifecycle policies", "D": "AWS CloudTrail"}, "answer": "C", "explanation": "DLM automates snapshot creation, retention, and cross-account sharing through lifecycle policies. S3 lifecycle policies manage objects, Config evaluates compliance, and CloudTrail audits API calls — none schedule EBS snapshots. ByteBar · SAA-C03 Practice Pack — for personal use only 28", "domain": "Domain 1"}
{"id": "bb-02058", "exam": "ISC2 CCSP", "exam_code": "CCSP", "product": "Practice Pack", "question": "Cloud Platform A platform team wants Kubernetes to automatically reject any pod whose container image is not signed by the corporate build pipeline. Which control enforces this at deploy time?", "options": {"A": "The kubelet agent", "B": "etcd encryption", "C": "A service mesh", "D": "Admission controllers"}, "answer": "D", "explanation": "Admission controllers (D) intercept API requests before objects are persisted and can enforce policies such as 'only images from the approved registry with a valid signature.' The kubelet (A) runs pods but does not set policy. etcd encryption (B) protects stored cluster state. A service mesh (C) manages service-to-service traffic, not deployment admission.", "domain": "Cloud Platform"}
{"id": "bb-02940", "exam": "CompTIA Cloud+", "exam_code": "CV0-004", "product": "Practice Pack", "question": "A team defines its entire cloud environment — networks, VMs, and storage — in version-controlled text files that can be deployed identically every time. This practice is called:", "options": {"A": "Infrastructure as code", "B": "Technical debt", "C": "Configuration drift", "D": "Shadow IT"}, "answer": "A", "explanation": "Infrastructure as code (IaC) defines infrastructure in machine-readable files, giving repeatable, versioned, reviewable deployments. Technical debt is accumulated shortcut cost, configuration drift is environments diverging over time, and shadow IT is unsanctioned technology use.", "domain": "Domain 2"}
{"id": "bb-02765", "exam": "AWS Certified Cloud Practitioner", "exam_code": "CLF-C02", "product": "Study Guide", "question": "An auditor wants continuous, automated collection of evidence showing that the company’s AWS usage meets its compliance requirements. Which service is built for this?", "options": {"A": "AWS Audit Manager", "B": "AWS Artifact", "C": "AWS Config", "D": "AWS CloudTrail"}, "answer": "A", "explanation": "Audit Manager continuously gathers evidence and produces audit-ready reports mapped to compliance frameworks. Artifact is where you download AWS’s own compliance reports; Config tracks resource configuration compliance; CloudTrail logs API activity — both feed evidence, but neither generates the audit report.", "domain": null}
{"id": "bb-00109", "exam": "Google Professional ML Engineer", "exam_code": null, "product": "Study Guide", "question": "A credit-risk team has tabular applicant data (income, history, demographics). Which model family should they try first?", "options": {"A": "A transformer fine-tuned on the text fields", "B": "ARIMA_PLUS on the application timestamps", "C": "Gradient boosted trees on the tabular features", "D": "K-means on the combined feature set"}, "answer": "C", "explanation": "For tabular data with mixed numeric/categorical features, gradient boosted trees are the strongest default — they handle non-linearities and missing values with minimal tuning. A transformer (A) is overkill for tabular rows and needs far more data. K-means (D) is unsupervised, not a classifier. ARIMA_PLUS (B) is time-series forecasting. Tabular classification = boosted trees first.", "domain": null}
{"id": "bb-01567", "exam": "Microsoft Azure Fundamentals", "exam_code": "AZ-900", "product": "Study Guide", "question": "[Domain 1] — Every Azure region is paired with another region at least 300 miles away. The PRIMARY purpose of region pairs is to:", "options": {"A": "Reduce latency for users near the region", "B": "Apply platform updates to both regions at the same time", "C": "Enable disaster recovery by failing over to the paired region", "D": "Provide free data transfer between any two regions"}, "answer": "C", "explanation": "Region pairs exist for disaster recovery and high availability: if one region suffers a major outage, services can fail over to its pair. Platform updates are deliberately rolled out to one region at a time, not simultaneously.", "domain": null}
{"id": "bb-04138", "exam": "PMHNP-BC", "exam_code": "PMHNP-BC", "product": "Study Guide", "question": "A patient with psychotic depression has failed three adequate antidepressant trials. Which treatment has the strongest evidence?", "options": {"A": "Electroconvulsive therapy (ECT)", "B": "St. John's wort", "C": "Sleep deprivation therapy", "D": "A fourth antidepressant monotherapy trial"}, "answer": "A", "explanation": "ECT is the gold standard for treatment-resistant and psychotic depression — the highest remission rates of any modality, and first-line for catatonia and severe suicidality. There are no absolute contraindications (increased intracranial pressure is the main caution). Memory loss is the signature side effect. A fourth monotherapy trial (D) after three failures has dismal odds.", "domain": null}
{"id": "bb-05642", "exam": "CompTIA Linux+", "exam_code": "XK0-006", "product": "Study Guide", "question": "Which command prints the UUID, label, and filesystem type of all block devices?", "options": {"A": "blkid", "B": "fdisk -l", "C": "df -T", "D": "mount"}, "answer": "A", "explanation": "blkid reports UUIDs, labels, and filesystem types from the block devices. fdisk -l shows partition tables; df -T covers only mounted filesystems; mount shows the mount table. ByteBar · CompTIA Linux+ (XK0-006) Study Guide — for personal use only 34", "domain": null}
{"id": "bb-05621", "exam": "CompTIA Linux+", "exam_code": "XK0-006", "product": "Study Guide", "question": "On Ubuntu, which tool installs a local .deb file AND resolves its dependencies?", "options": {"A": "dpkg -i", "B": "apt install ./package.deb", "C": "rpm -i", "D": "snap install"}, "answer": "B", "explanation": "dpkg -i installs but does NOT resolve dependencies — the classic trap. apt (even on a local .deb path) fetches and resolves them. rpm is for RPM-based distros.", "domain": null}
{"id": "bb-05113", "exam": "CompTIA Security+", "exam_code": "SY0-701", "product": "Practice Pack", "question": "A NIPS detects an exploit attempt and drops the packets automatically. The NIPS is acting as a:", "options": {"A": "Detective control", "B": "Preventive control", "C": "Deterrent control", "D": "Directive control"}, "answer": "B", "explanation": "Inline blocking = preventive. A NIDS that only alerts would be detective — the -P- is the difference.", "domain": null}
{"id": "bb-04661", "exam": "AWS Solutions Architect Associate", "exam_code": "SAA-C03", "product": "Practice Pack", "question": "An analytics job scans 500 GB CSV files in S3 but only needs 5% of the columns. How can data transfer and cost be reduced?", "options": {"A": "Download the full files faster with Transfer Acceleration", "B": "Replicate the files to more Regions", "C": "Use S3 Select to retrieve only the needed columns server-side", "D": "Increase the EC2 instance size running the job"}, "answer": "C", "explanation": "S3 Select filters data inside S3 using SQL expressions, returning only the needed columns and slashing transfer. Transfer Acceleration speeds uploads, replication multiplies storage, and a bigger instance doesn't reduce the data scanned.", "domain": "Domain 2"}
{"id": "bb-00351", "exam": "Cisco CCNA", "exam_code": "200-301", "product": "Practice Pack", "question": "What is 2001:db8::/32 in IPv6?", "options": {"A": "Link-local range", "B": "Documentation range", "C": "Multicast range", "D": "Loopback"}, "answer": "B", "explanation": "2001:db8::/32 is reserved for documentation/examples — never routed on the internet.", "domain": "D1"}
{"id": "bb-05423", "exam": "TEAS 7", "exam_code": "TEAS-7", "product": "Study Guide", "question": "Find the median of: 7, 3, 9, 3, 8.", "options": {"A": "3", "B": "6", "C": "7", "D": "9"}, "answer": "C", "explanation": "Sorted: 3, 3, 7, 8, 9. The middle value is 7.", "domain": null}
{"id": "bb-05618", "exam": "CompTIA Linux+", "exam_code": "XK0-006", "product": "Study Guide", "question": "Key-based SSH logins fail with 'bad ownership or modes' for the key files. Which permissions fix ~/.ssh and ~/.ssh/authorized_keys?", "options": {"A": "chmod 777 ~/.ssh and chmod 666 ~/.ssh/authorized_keys", "B": "chown root:root ~/.ssh/authorized_keys", "C": "chmod 700 ~/.ssh and chmod 600 ~/.ssh/authorized_keys", "D": "chmod 755 ~/.ssh and chmod 644 ~/.ssh/authorized_keys"}, "answer": "C", "explanation": "sshd StrictModes requires 700 on ~/.ssh and 600 on authorized_keys; anything group- or world-accessible is rejected. Giving them to root breaks the user's ownership.", "domain": null}
{"id": "bb-05302", "exam": "TEAS 7", "exam_code": "TEAS-7", "product": "Practice Pack", "question": "A researcher tests whether a new wound dressing speeds healing. She applies the new dressing to one group of patients and the standard dressing to another, then compares healing times. The type of dressing used is the:", "options": {"A": "Dependent variable, because it depends on the healing time.", "B": "Control group, because it is being compared.", "C": "Constant, because it stays the same for everyone.", "D": "Independent variable, because the researcher manipulates it."}, "answer": "D", "explanation": "D is correct because the independent variable is the factor the researcher deliberately changes — here, the dressing type assigned to each group. A is wrong because the dependent variable is the measured outcome (healing time), not the treatment; dressing type does not 'depend on' healing time. B is wrong because the control group is the group of patients receiving the standard dressing, not the variable itself. C is wrong because a constant is held the same across groups; the dressing type is precisely what differs between the groups.", "domain": null}
{"id": "bb-04472", "exam": "Red Hat RHCSA/RHCE", "exam_code": "RHCSA", "product": "Practice Pack", "question": "Running grep -r \"PermitRootLogin\" * inside /etc/ssh finds nothing, yet grep -r \"PermitRootLogin\" /etc/ssh finds matches. What explains this?", "options": {"A": "The shell expands * before grep runs, and * omits dotfiles and dot-directories; passing the directory path itself avoids the glob and searches everything", "B": "grep -r cannot descend into subdirectories, so only the top level is ever searched", "C": "Hidden files are encrypted on RHEL and cannot be searched by grep at all", "D": "grep -R (capital R) is required for any recursive search; lowercase -r only searches one level"}, "answer": "A", "explanation": "itself avoids the glob and searches everything Glob expansion is done by the shell, not grep: * expands to all non-dot entries in the current directory, so dotfiles and dot-directories are never handed to grep. Giving grep the literal path /etc/ssh lets grep's own recursion walk the entire tree including dot entries. Option B is false; -r is exactly the recursive flag. Option D confuses -R, which additionally follows symlinked directories during recursion. Option C invents a nonexistent encryption behavior.", "domain": "Essential Tools & Running Systems"}
{"id": "bb-05159", "exam": "CompTIA Security+", "exam_code": "SY0-701", "product": "Practice Pack", "question": "Before running an intrusive vulnerability scan against production servers, the team must FIRST:", "options": {"A": "Get written authorization and schedule it in a maintenance window", "B": "Run it at noon for realism", "C": "Disable logging", "D": "Scan from home"}, "answer": "A", "explanation": "Intrusive scans can crash systems — authorization + change control + off-hours window. Scanning without permission is indistinguishable from attacking.", "domain": null}
{"id": "bb-02805", "exam": "CompTIA CySA+", "exam_code": "CS0-004", "product": "Practice Pack", "question": "A threat intelligence feed provides an indicator in STIX format. Your SIEM does not natively parse STIX. Which protocol is designed to let you automatically retrieve such shared threat intelligence?", "options": {"A": "NetBIOS", "B": "RDP", "C": "TAXII", "D": "SNMP"}, "answer": "C", "explanation": "TAXII (Trusted Automated eXchange of Intelligence Information) is the transport protocol for sharing cyber threat intelligence, commonly carrying STIX-formatted indicators. SNMP is for device management, NetBIOS is legacy name resolution, and RDP is remote administration.", "domain": "Domain 1"}