85+ Vault Interview Questions by Exam Objective (2027)

88 HashiCorp Vault interview questions tagged to Vault Associate exam objectives: auth, policies, tokens, leases, engines, transit, architecture, VSO.

Sep 25, 2025 - 10:39
Nov 6, 2026 - 21:25
 0  66
85+ Vault Interview Questions by Exam Objective (2027)

HashiCorp Vault interviews and the Vault Associate exam test the same nine areas: auth methods, policies, tokens, leases, secrets engines, encryption as a service, architecture, deployment and access management. This page gives 88 questions with direct answers, each tagged with the exam objective it maps to, so you can prepare for both at once.

Objectives follow HashiCorp's published exam content list for Vault Associate (003). HashiCorp's certification page currently lists Vault 1.19 as the tested version, so check the official exam page for any updated objective list before you book. If you want story style questions instead, use our scenario-based Vault interview questions, the main page in this series. Beginners should start with most asked Vault questions for freshers.

Key takeaways

  • Every question below carries an objective tag such as (1a) or (5c), matching the Vault Associate exam content list.
  • Freshers are tested on objectives 1 to 4: auth methods, policies, tokens and leases.
  • Candidates with 2 to 5 years of experience get secrets engines, transit and architecture questions (objectives 5 to 7).
  • Senior candidates get deployment architecture, replication, Vault Agent and the Vault Secrets Operator (objectives 8 and 9), plus operations.
  • Scenario questions at the end combine several objectives, which is how real interviews work.

Exam objectives at a glance

Objective areaWhat it coversQuestions here
1. Authentication methodsPurpose, choosing, human vs system, identities, configuringQ1 to Q10
2. Vault policiesPath, capabilities, choosing and writing policiesQ11 to Q19
3. Vault tokensService vs batch, root, accessors, TTL, orphansQ20 to Q29
4. Vault leasesLease IDs, renew, revokeQ30 to Q34
5. Secrets enginesDynamic vs static, response wrapping, enablingQ35 to Q48
6. Encryption as a serviceTransit encrypt, decrypt, rotateQ49 to Q54
7. Architecture fundamentalsBarrier, seal and unseal, environment variablesQ55 to Q61
8. Deployment architectureStorage, Shamir, replication, self-managed vs HCPQ62 to Q70
9. Access management architectureVault Agent, Vault Secrets OperatorQ71 to Q76
ScenariosMixed objectivesQ77 to Q88

Freshers: objectives 1 to 4

Q1. What is the purpose of an auth method in Vault? (1a)

An auth method verifies who a client is and returns a token with policies attached. Vault does not care how you proved identity once the token exists; every later request uses that token. Examples include userpass, LDAP, OIDC, AppRole, Kubernetes and AWS.

Q2. Which auth method is always enabled? (1a)

The token auth method. It is mounted at auth/token/, cannot be disabled, and is what every other auth method ultimately produces.

Q3. How do you enable an auth method from the CLI? (1f)

Run vault auth enable <type>, for example vault auth enable approle. Add -path= to mount it at a custom path, which lets you run two LDAP backends side by side.

Q4. What is the difference between human and system auth methods? (1c)

Human methods (userpass, LDAP, OIDC, GitHub) suit people who can type a password or complete a browser login. System methods (AppRole, Kubernetes, AWS, Azure, GCP, TLS certificates) suit machines that prove identity with a platform credential or a role ID and secret ID.

Q5. Which auth method would you choose for a pod in Kubernetes? (1b)

The Kubernetes auth method. The pod sends its service account token, Vault validates it with the Kubernetes TokenReview API, and a role maps the service account and namespace to policies.

Q6. Which auth method fits an EC2 instance or Lambda function? (1b)

The AWS auth method. The IAM type lets the workload sign a request with its IAM credentials, so there is no secret to distribute.

Q7. What is AppRole and when do you use it? (1b)

AppRole is a machine auth method that uses a role ID (like a username) and a secret ID (like a password). Use it for CI systems or servers where no platform identity method fits, and deliver the secret ID through a separate, trusted path.

Q8. What are entities and groups? (1d)

An entity represents one person or service across auth methods; each login maps to an alias on the entity. Groups collect entities, and policies attached to an entity or group apply no matter which auth method the user logged in with.

Q9. How do you log in from the CLI with userpass? (1e)

vault login -method=userpass username=alice. The CLI prompts for the password and stores the returned token in the token helper (by default ~/.vault-token).

Q10. How do you authenticate with the HTTP API? (1e)

Send a POST to the auth method's login path, for example /v1/auth/userpass/login/alice with the password in the JSON body. The response contains auth.client_token, which you send in the X-Vault-Token header on later calls.

Q11. Why does Vault need policies? (2a)

Vault denies everything by default. Policies grant access to paths, so they are how you enforce least privilege. A token can only do what its attached policies allow.

Q12. What does the path element in a policy do? (2b)

It names the API path the rule applies to. A trailing * is a prefix glob, and + matches a single path segment, so secret/data/+/config matches any one team's config.

Q13. List the policy capabilities. (2c)

create, read, update, patch, delete, list, sudo and deny. sudo is needed for root protected paths, and deny overrides every other capability.

Q14. Write a policy that lets an app read its KV v2 secrets. (2d)

path "secret/data/payments/*" {
  capabilities = ["read"]
}

KV version 2 inserts data/ into the API path, which is a common exam trap. Listing keys needs list on secret/metadata/payments/*.

Q15. What happens if two policies conflict on the same path? (2c)

Capabilities are merged, except deny, which always wins. For different paths, the most specific matching path is used.

Q16. Which built-in policies exist? (2a)

default, attached to every token unless you opt out, which lets a token look up and renew itself; and root, which allows everything and cannot be changed.

Q17. How do you upload a policy from the CLI? (2e)

vault policy write payments-read payments-read.hcl. Use vault policy read to view it and vault policy list to list all policies.

Q18. How can you test what a token is allowed to do? (2d)

vault token capabilities <token> secret/data/payments/db returns the capabilities that token has on that path.

Q19. Can policies be created from the UI? (2e)

Yes. The UI has a Policies section where you can create and edit ACL policies with the same HCL syntax.

Q20. What are service tokens and batch tokens? (3a)

Service tokens are persisted, renewable, can create child tokens and have accessors. Batch tokens are encrypted blobs that are not stored, cannot be renewed, revoked manually or create children, and are cheap to issue, which suits high volume, short lived workloads.

Q21. How can you tell them apart by prefix? (3a)

In current Vault versions, service tokens start with hvs. and batch tokens with hvb.. Recovery tokens use hvr..

Q22. When should the root token be used? (3b)

Only for initial setup and emergencies. Create admin policies and auth methods, then revoke the root token with vault token revoke. Generate a new one later with vault operator generate-root and a quorum of unseal or recovery key holders.

Q23. What is a token accessor? (3c)

A reference to a token that lets you look up, renew or revoke it without knowing the token itself. Audit logs and admin tooling use accessors so the real token is never exposed.

Q24. What does TTL mean for a token? (3d)

Time to live is how long the token is valid before it expires. Tokens can be renewed up to their max TTL. When the token expires, its child tokens and the leases it created are revoked too.

Q25. What is the default system max TTL? (3d)

The system max TTL is 32 days (768 hours) unless you change it in the server configuration. Mount tuning, roles and individual tokens can set shorter values.

Q26. What is an orphan token? (3e)

A token with no parent. Normally, revoking a parent revokes its children; an orphan survives that. Create one with vault token create -orphan (needs sudo) or through a role that issues orphans.

Q27. What is a periodic token? (3f)

A token with a period instead of a max TTL. As long as it is renewed within each period, it never expires, which suits long running services. Creating one requires sudo or a token role.

Q28. How do you create a token with a specific policy and TTL? (3f)

vault token create -policy=payments-read -ttl=1h. Add -use-limit= to cap how many requests it can make.

Q29. How do you look up details of your current token? (3c)

vault token lookup shows its policies, TTL, accessor and whether it is renewable or orphaned.

Q30. What is a lease? (4a)

Metadata attached to dynamic secrets and some auth results: a lease ID, a duration and whether it is renewable. When the lease expires or is revoked, Vault revokes the underlying credential, for example dropping a database user.

Q31. What does a lease ID look like? (4a)

It starts with the path that created it, for example database/creds/readonly/abc123.... That prefix is what lets you revoke all leases from one path at once.

Q32. How do you renew a lease? (4b)

vault lease renew <lease_id>, optionally with -increment=. Renewal cannot extend past the max TTL.

Q33. How do you revoke a lease, or all leases under a path? (4c)

vault lease revoke <lease_id> revokes one. vault lease revoke -prefix database/creds/readonly revokes every lease under that prefix, which is the fast response to a leaked credential.

Q34. Do static KV secrets have leases? (4a)

No. KV secrets are stored values without a lease to revoke. KV v1 returns a lease_duration as a hint for how often to refresh, but nothing is revoked when it passes.

2 to 5 years: objectives 5 to 7

Q35. What is a secrets engine? (5d)

A component mounted at a path that stores, generates or encrypts data. Examples: KV stores static secrets, database and AWS generate credentials, PKI issues certificates and transit encrypts data.

Q36. How do you enable a secrets engine with the CLI, API and UI? (5g)

CLI: vault secrets enable -path=kv kv-v2. API: POST to /v1/sys/mounts/kv with {"type":"kv","options":{"version":"2"}}. UI: Secrets Engines, then Enable new engine.

Q37. Static vs dynamic secrets? (5b)

Static secrets are values you store and rotate yourself, such as an API key in KV. Dynamic secrets are generated on request with a lease, unique per client, and revoked automatically, such as a database user created by the database engine.

Q38. Why are short lived dynamic secrets valuable? (5f)

A leaked credential expires on its own, every client has a separate credential so you can trace and revoke one, and nobody has to rotate a shared password.

Q39. Which engine would you choose for temporary AWS access? (5a)

The AWS secrets engine. It can create IAM users, assume roles or issue federation tokens and return short lived credentials scoped by role.

Q40. Which engine for an internal certificate authority? (5a)

The PKI engine. It can act as a root or intermediate CA and issue short lived certificates through roles, which reduces the need for revocation lists.

Q41. What are the differences between KV v1 and v2? (5a)

KV v2 keeps versions of each secret, supports soft delete, undelete and destroy, and has metadata with check-and-set. KV v1 stores only the latest value and is slightly faster.

Q42. How do you read a specific version of a KV v2 secret? (5h)

vault kv get -version=3 secret/payments/db. vault kv rollback -version=3 writes that version back as the newest.

Q43. How do you read a secret through the API? (5h)

GET /v1/secret/data/payments/db with the X-Vault-Token header. The value is under data.data for KV v2.

Q44. What is response wrapping? (5e)

Vault places the real response in a cubbyhole and returns a single use wrapping token with a short TTL. The receiver unwraps it with vault unwrap. If someone intercepted and used it first, the unwrap fails, so you know.

Q45. How do you request a wrapped response? (5e)

Add -wrap-ttl= to the CLI command, for example vault write -wrap-ttl=60s -f auth/approle/role/ci/secret-id, or send the X-Vault-Wrap-TTL header with an API request.

Q46. What is the cubbyhole engine? (5d)

A per-token private storage area enabled by default. Only the token that wrote the data can read it, and it is destroyed when the token expires. Response wrapping is built on it.

Q47. How does the database secrets engine create credentials? (5b)

You configure a connection with a privileged user and define a role with creation SQL. Each read of database/creds/<role> runs that SQL to create a new user with a lease, and revocation drops the user.

Q48. What are static roles in the database engine? (5b)

Static roles map to an existing database user and have Vault rotate its password on a schedule. Use them for accounts that cannot be recreated, such as a legacy app user.

Q49. What problem does the transit engine solve? (5c)

It lets applications encrypt and decrypt data without handling keys. Vault keeps the keys and does not store the data; the application stores the ciphertext.

Q50. How do you encrypt data with transit? (6a)

vault secrets enable transit
vault write -f transit/keys/orders
vault write transit/encrypt/orders plaintext=$(echo "secret" | base64)

Plaintext must be base64 encoded. The ciphertext starts with vault:v1:.

Q51. How do you decrypt? (6a)

vault write transit/decrypt/orders ciphertext="vault:v1:..." returns base64 plaintext, which you decode with base64 --decode.

Q52. How do you rotate a transit key, and what happens to old data? (6b)

vault write -f transit/keys/orders/rotate adds a new key version. New encryptions use it; old ciphertext still decrypts with its recorded version.

Q53. What is rewrap? (6b)

transit/rewrap/<key> re-encrypts ciphertext with the latest key version without returning plaintext to the caller. Use it after rotation, then raise min_decryption_version to retire old versions.

Q54. What is the default transit key type? (6a)

aes256-gcm96. Transit also supports other types, such as ChaCha20-Poly1305, RSA and ECDSA, for different uses including signing.

Q55. How does Vault encrypt its own data? (7a)

Everything written to storage passes through the cryptographic barrier and is encrypted with the encryption key from the keyring. The keyring is protected by the root key, which is protected by the unseal keys or an auto-unseal mechanism.

Q56. What does sealed mean? (7b)

Vault knows where its data is but cannot decrypt it, because the root key is not in memory. A sealed Vault answers only status and unseal requests.

Q57. How do you unseal Vault with Shamir keys? (7b)

Run vault operator unseal once per key holder until the threshold is reached, for example 3 of 5. Each node in a cluster must be unsealed.

Q58. What is auto-unseal? (7b)

Vault uses a trusted external service, such as AWS KMS, Azure Key Vault, Google Cloud KMS or an HSM, to decrypt the root key at startup. Admins then hold recovery keys instead of unseal keys.

Q59. How do you seal Vault in an emergency? (7b)

vault operator seal, which needs sudo on sys/seal. Sealing discards the root key from memory and stops all secret access until unsealed.

Q60. Which environment variables does the CLI use? (7c)

VAULT_ADDR for the server address, VAULT_TOKEN for the token, VAULT_NAMESPACE for Enterprise and HCP namespaces, and VAULT_CACERT or VAULT_SKIP_VERIFY for TLS (never skip verification in production).

Q61. What does vault status show? (7b)

Seal status, seal type, key shares and threshold, version, storage type, HA mode and whether this node is active or standby.

Senior: objectives 8 and 9, plus operations

Q62. Which storage backend is recommended? (8b)

Integrated Storage, which uses the Raft consensus protocol and needs no external system. Consul storage is still supported, but Integrated Storage is the default choice for new clusters.

Q63. How many nodes should an Integrated Storage cluster have? (8a)

An odd number, typically 3 or 5, so Raft can keep a quorum. Five nodes tolerate two failures; three tolerate one.

Q64. Why use Shamir's secret sharing? (8c)

It splits the unseal key into shares so no single person can unseal Vault. A threshold of holders must cooperate, which protects against a rogue admin.

Q65. What is the difference between DR and performance replication? (8d)

Both are Enterprise features. DR replication copies everything, including tokens and leases, to a standby cluster that does not serve clients until promoted. Performance replication copies configuration and secrets to secondaries that serve reads locally, but tokens and leases stay local.

Q66. When would you use performance standby nodes? (8a)

In Enterprise clusters, performance standbys serve read requests that do not change state, spreading load away from the active node. Writes are still forwarded to the active node.

Q67. Self-managed Vault vs HCP Vault Dedicated? (8e)

Self-managed means you run the servers, storage, upgrades and unseal. HCP Vault Dedicated is HashiCorp's managed service, where HashiCorp operates the cluster and you manage policies, auth and engines.

Q68. How do you take a snapshot of Integrated Storage? (8b)

vault operator raft snapshot save backup.snap, and restore with vault operator raft snapshot restore. Enterprise adds automated snapshots.

Q69. How do you check Raft cluster members? (8a)

vault operator raft list-peers shows node IDs, addresses, the leader and voter status.

Q70. What is a namespace? (8a)

An isolated tenant in Vault Enterprise and HCP Vault with its own auth methods, secrets engines, policies and tokens. Use namespaces to give teams self-service without sharing a policy space.

Q71. What is Vault Agent? (9a)

A client side daemon that authenticates with an auto-auth method, keeps the token renewed, caches responses, and renders secrets into files with templates. Applications read a file instead of calling Vault.

Q72. What does auto-auth do in Vault Agent? (9a)

It logs in using a configured method (Kubernetes, AWS, AppRole and others), writes the token to sinks, and re-authenticates when the token expires.

Q73. What is the Vault Secrets Operator? (9b)

A Kubernetes operator that syncs secrets from Vault into native Kubernetes Secrets using custom resources such as VaultStaticSecret and VaultDynamicSecret. Apps consume normal Secrets, and the operator refreshes them and can trigger a rollout restart of workloads when a secret rotates.

Q74. Vault Agent injector or Vault Secrets Operator? (9b)

The Agent injector adds a sidecar that writes secrets to a shared volume, so secrets never become Kubernetes Secrets. The Secrets Operator creates Kubernetes Secrets, which is simpler for apps but means etcd encryption must be on.

Q75. What is Vault Proxy? (9a)

A daemon split out from Agent's API proxy features. It forwards application requests to Vault with an auto-auth token and can cache them, without templating.

Q76. How do you enable audit logging? (operations)

vault audit enable file file_path=/var/log/vault_audit.log. Enable at least two audit devices in production: if all enabled devices fail to write, Vault refuses requests rather than serve them unaudited.

Scenario questions

Q77. A CI pipeline stores a long lived database password in its variables. How do you fix it? (1b, 5b, 5f)

Give the pipeline a platform identity (JWT/OIDC auth for GitHub Actions or GitLab, or AppRole) and a policy that only reads database/creds/ci. Each run gets a short lived user that is revoked at lease end. Remove the stored password and rotate it.

Q78. A developer leaked a dynamic database credential in a log. What do you do? (4c)

Revoke that lease by ID, or all leases under the role with vault lease revoke -prefix if you cannot identify it. Use the audit log (HMACed values plus the accessor) to see who requested it, then tighten the role's TTL.

Q79. A policy allows secret/payments/* but the app gets 403 on KV v2. Why? (2b, 5h)

KV v2 paths include data/, so the policy must target secret/data/payments/*, plus secret/metadata/payments/* for list. The CLI hides the prefix, the policy does not.

Q80. After a reboot, all Vault nodes are sealed and two key holders are on leave. Options? (7b, 8c)

With Shamir you need the threshold of holders; there is no bypass. Long term, move to auto-unseal with a cloud KMS or HSM so restarts need no humans, and keep recovery keys for privileged operations.

Q81. A batch job requests thousands of tokens per minute and storage is growing. What do you change? (3a)

Use batch tokens for that role. They are not persisted, so they do not add storage writes, and the job does not need renewal or child tokens.

Q82. Security wants field level encryption for card tokens in PostgreSQL without the app managing keys. Design it. (5c, 6a, 6b)

Create a transit key per application, let the app encrypt and decrypt through a policy limited to that key, store the ciphertext in the column, rotate on a schedule, rewrap in a background job and raise min_decryption_version afterwards.

Q83. A team needs a separate Vault admin space without seeing other teams' secrets. (8a, 2d)

On Enterprise or HCP, create a namespace per team and delegate admin inside it. On Community Edition, use path prefixes per team, team specific policies and identity groups, and keep sys/ admin rights central.

Q84. Pods must read secrets but the security team forbids Kubernetes Secrets. Which pattern? (9a, 9b)

Use the Vault Agent injector or the Vault CSI provider so secrets are written to an in-memory volume. The Vault Secrets Operator would create Kubernetes Secrets, which breaks the rule.

Q85. The primary region fails. How do you recover service? (8d)

With DR replication, promote the DR secondary, point DNS or the load balancer at it, and let clients reconnect with their existing tokens, which were replicated. Without Enterprise, restore the latest Raft snapshot into a new cluster and expect clients to re-authenticate.

Q86. An auditor asks who read a secret last Tuesday. Where do you look? (operations)

The audit device logs. Request paths are in clear, while tokens and secret values are HMACed. Use sys/audit-hash to hash a known token or accessor and search for it.

Q87. A root token was found in a shell history file. Response? (3b)

Revoke it immediately, check the audit log for its use, review every change made with it, and generate future root tokens only with vault operator generate-root and a quorum, revoking them after use.

Q88. Certificates for internal services keep expiring unnoticed. How can Vault help? (5a, 9a)

Use the PKI engine with short TTLs and let Vault Agent templates or cert-manager with a Vault issuer renew certificates automatically, so expiry becomes routine instead of an incident.

How to use this list

Vault is a core part of our DevSecOps training, with labs on auth methods, dynamic secrets, transit and Kubernetes integration, in Kharadi, Pune and live online.

Frequently asked questions

How many questions are on this page?

88 numbered questions: 34 for freshers, 27 for 2 to 5 years, 15 for senior roles and 12 scenarios, each tagged with a Vault Associate objective.

Which Vault version does the Associate exam test?

HashiCorp's certification page lists Vault 1.19 for the Vault Associate exam. Check the official exam page before booking, because the tested version and objective list can change.

What is the hardest Vault Associate objective?

Most candidates find tokens (objective 3) hardest, especially orphan, periodic and batch tokens, and how revocation cascades from parent to child tokens.

Do I need Vault Enterprise to prepare?

No. The dev server in the free Vault binary covers almost every objective. Replication and namespaces are Enterprise features, so learn them from the documentation.

What is the difference between Vault Agent and the Vault Secrets Operator?

Vault Agent authenticates and renders secrets to files for one application. The Vault Secrets Operator syncs Vault secrets into Kubernetes Secrets for any pod in the cluster.

Is the transit engine on the exam?

Yes. Objective 5c covers when to use transit, and objective 6 covers encrypting, decrypting and rotating keys.

References

What's Your Reaction?

Like Like 0
Dislike Dislike 0
Love Love 0
Funny Funny 0
Angry Angry 0
Sad Sad 0
Wow Wow 0
Mridul I am a passionate technology enthusiast with a strong focus on DevOps, Cloud Computing, and Cybersecurity. Through my blogs at DevOps Training Institute, I aim to simplify complex concepts and share practical insights for learners and professionals. My goal is to empower readers with knowledge, hands-on tips, and industry best practices to stay ahead in the ever-evolving world of DevOps.