80+ JFrog Interview Questions and Answers (2027)

82 JFrog interview questions with answers: Artifactory repositories, build info, Xray, Curation, release bundles, JFrog CLI, HA, plus scenarios by level.

Sep 24, 2025 - 16:15
Nov 8, 2026 - 04:06
 0  70
80+ JFrog Interview Questions and Answers (2027)

JFrog interviews test whether you can run an artifact platform, not only push files to it. Expect questions on Artifactory repository types, build info, permissions and tokens, Xray watches and policies, Curation, release bundles and distribution, the JFrog CLI, and how to keep the platform fast and available. This page has 82 questions with direct answers, grouped by experience level, plus scenarios.

This is the full question bank in our JFrog series. For definitions only, start with most asked JFrog questions for freshers. For exam style preparation, use JFrog certification questions, and for longer production stories, read scenario-based JFrog questions.

Key takeaways

  • Know the four repository types (local, remote, virtual, federated) and the resolution order of a virtual repository.
  • Build info is what links an artifact to its source commit, dependencies and environment. Most senior questions build on it.
  • Xray scans artifacts and builds through watches and policies; Curation blocks risky open source packages before they are cached.
  • Release Bundles v2 group approved artifacts for promotion and distribution.
  • JFrog Pipelines reached end of life on 1 May 2026, and Mission Control now runs as a microservice inside Artifactory. Answer questions about them in that light.
  • Use access tokens, not passwords or legacy API keys, in pipelines.

JFrog platform components

ComponentWhat it doesTypical interview angle
ArtifactoryUniversal binary repository for Maven, npm, PyPI, Docker, Helm, Go, Terraform and moreRepository design, storage, permissions
XraySoftware composition analysis: vulnerabilities, licenses, operational riskWatches, policies, blocking downloads
CurationBlocks malicious or non-compliant open source packages at the remote repositoryShift left supply chain control
DistributionMoves signed release bundles to edge nodes and other sitesRelease flow, multi-site
JFrog CLI (jf)Command line for upload, download, build info, scanning and release bundlesCI integration

Freshers (0 to 2 years)

Q1. What is JFrog Artifactory?

A binary repository manager. It stores build outputs and caches third party dependencies for many package types in one place, with metadata, permissions and search.

Q2. Why not store binaries in Git?

Git is built for text diffs. Large binaries bloat the repository, cannot be deduplicated well and do not carry build metadata. A repository manager stores each binary once, serves it through native package protocols and keeps an audit trail.

Q3. What is a local repository?

A repository that stores artifacts your organization uploads or builds, such as release JARs or internal Docker images.

Q4. What is a remote repository?

A caching proxy for an external registry such as Maven Central, npmjs or Docker Hub. The first request downloads and caches the package; later requests are served from the cache.

Q5. What is a virtual repository?

A single URL that aggregates local and remote repositories. Clients point at one address and Artifactory resolves from the members.

Q6. In what order does a virtual repository resolve artifacts?

Local repositories first, then remote repository caches, then the remote repositories themselves, following the order you configure among members.

Q7. What is a federated repository?

A repository that stays in sync with matching repositories on other JFrog Platform Deployments. Each federated repository can have up to 10 members, and it supports all package types except TerraformBackend.

Q8. Name package types Artifactory supports.

Maven, Gradle, npm, PyPI, NuGet, Go, Docker, Helm, Conan, RubyGems, Debian, RPM, Terraform and generic, among others.

Q9. What is a generic repository?

A repository with no package protocol, used for files such as zip archives, installers or reports.

Q10. What is checksum based storage?

Artifactory stores each binary once, keyed by its checksum, and keeps paths as database records pointing to it. The same file in ten repositories uses disk space once, and copying is a metadata operation.

Q11. What are properties in Artifactory?

Key value metadata attached to files or folders, such as build.number=142 or qa.status=passed. Searches, cleanup and promotion rules can use them.

Q12. How do you upload a file with the JFrog CLI?

jf rt upload "target/*.jar" libs-release-local/com/acme/app/1.4.0/. The short alias is jf rt u.

Q13. How do you configure the CLI to talk to a server?

jf config add stores a server ID with its URL and credentials. jf rt ping checks connectivity.

Q14. How do you point Maven at Artifactory?

Set a mirror or repository in settings.xml to the virtual repository URL, or use jf mvn-config and run builds through jf mvn, which also records build info.

Q15. How do you push a Docker image to Artifactory?

Log in to the Artifactory Docker registry, tag the image with the registry host and repository, and push it. The JFrog CLI can wrap the push with jf docker push to collect build info.

Q16. What is JFrog Xray?

A software composition analysis tool integrated with Artifactory. It indexes artifacts and builds, finds known vulnerabilities and license issues in their components, and can block downloads that violate policy.

Q17. What is a repository layout?

A pattern that tells Artifactory how paths map to organization, module, version and classifier. Layouts make version based cleanup and searches possible for non Maven content.

Q18. What is the Set Me Up feature?

A UI helper that shows the client configuration snippet for a repository, for example the npm registry URL or Maven settings, so users can connect without guessing.

Q19. Which ports does Artifactory 7 use?

The Router listens on 8082 for the UI, Access and platform APIs, and the Artifactory service listens on 8081. Internal services use more ports; check the network ports page for your version.

Q20. What is a SNAPSHOT repository in Maven terms?

A local repository that accepts changing versions ending in -SNAPSHOT, kept apart from release repositories so release versions stay immutable.

Q21. What is a smart remote repository?

A remote repository that points to another Artifactory instance. It can show remote browsing and sync properties and deletes, which suits branch office setups.

Q22. How do you download artifacts with the CLI?

jf rt download libs-release-local/com/acme/app/1.4.0/ ./out/, alias jf rt dl. Add --flat to drop the folder structure.

Q23. What is AQL?

Artifactory Query Language, a JSON-like query language for searching items, builds and their properties through the REST API or file specs.

Q24. What are file specs?

JSON files that describe upload or download patterns, targets, properties and AQL queries, so CI jobs use the same declarative definition every time.

2 to 5 years

Q25. What is build info?

A JSON record published with a build: modules, produced artifacts, resolved dependencies, environment variables, VCS revision and CI job. It lets you trace any binary back to its source and inputs.

Q26. How do you publish build info with the CLI?

jf rt upload "dist/*.tgz" npm-local/ --build-name=web --build-number=142
jf rt build-collect-env web 142
jf rt build-add-git web 142
jf rt build-publish web 142

--build-name and --build-number must always be passed together.

Q27. What is build promotion?

Moving or copying a build's artifacts to the next repository, for example from staging-local to release-local, and recording a status. jf rt build-promote does it from the CLI.

Q28. Copy or move during promotion?

Copy is cheap because of checksum storage and keeps history in the source repository. Move keeps the source clean. Many teams copy to production repositories and clean staging with retention rules.

Q29. How do you manage access in Artifactory?

Users and groups get permissions on repositories, builds and release bundles. Permission targets define resources and actions such as read, annotate, deploy, delete and manage. Grant to groups, never to individual users.

Q30. Which credential should a CI job use?

A scoped access token with only the permissions the job needs and an expiry. Avoid user passwords, and replace legacy API keys, which JFrog has deprecated in favour of access tokens.

Q31. How do you integrate Artifactory with SSO?

Configure SAML, OAuth or OpenID Connect providers, and LDAP or Crowd groups, so user and group membership comes from the identity provider instead of local accounts.

Q32. What are Xray watches?

A watch defines what to monitor (repositories, builds or release bundles) and which policies apply to it. Without a watch, scan results are visible but no policy action runs.

Q33. What policy types does Xray support?

Security policies (CVE severity or CVSS), license policies (allowed or banned licenses) and operational risk policies (such as outdated or unmaintained components).

Q34. What actions can an Xray policy trigger?

Generate a violation, notify by email or webhook, block downloads of the artifact, fail a build when the CI integration scans it, and others depending on the policy type.

Q35. What is the difference between jf audit and jf scan?

jf audit scans a source project's dependencies from its package manager files. jf scan scans files or binaries on disk. jf docker scan scans a local Docker image.

Q36. What is JFrog Curation?

A service that checks open source packages against policies when they are requested through a remote repository and blocks malicious, vulnerable or non compliant ones before they are cached. jf curation-audit shows which project dependencies would be blocked.

Q37. How do Xray and Curation differ?

Curation stops a package from entering your organization. Xray analyses what is already stored or built. You want both: one at the gate, one on everything inside.

Q38. What is a release bundle?

An immutable, signed collection of artifacts that make up a release. Release Bundles v2 are created from builds, other bundles or file specs and promoted through environments.

Q39. How do you create and promote a release bundle from the CLI?

jf release-bundle-create (alias jf rbc) creates it, and jf release-bundle-promote (alias jf rbp) promotes it to an environment such as QA or PROD.

Q40. What is JFrog Distribution?

The service that ships release bundles to edge nodes and other sites. Edge nodes are read-only Artifactory instances near consumers, such as regional data centres.

Q41. How do you clean up old artifacts?

Use Artifactory cleanup policies or retention scripts based on AQL (for example, delete snapshot artifacts not downloaded in 90 days and not part of a released build). Run garbage collection afterwards to free binary storage.

Q42. Why does deleting an artifact not free disk space at once?

Because the binary may be referenced by other paths. Space is reclaimed only when garbage collection finds the checksum has no references, and items in the trash can are still referenced until it is emptied.

Q43. What is the trash can?

A temporary repository that holds deleted items for a configured period so you can restore mistakes. Items there still occupy storage.

Q44. How do you set properties from the CLI?

jf rt set-props "release-local/app/1.4.0/*" "qa.status=passed", alias jf rt sp.

Q45. How does Artifactory handle Docker image layers?

Each layer is a blob stored once by checksum, so images that share base layers share storage. Manifests and tags are paths that reference those blobs.

Q46. What is a Docker remote repository used for?

Caching Docker Hub or other registries. It reduces pull rate limits and outages from the upstream registry, and Xray can scan cached images.

Q47. How do you integrate Artifactory with Jenkins?

Use the Jenkins JFrog plugin, which installs the JFrog CLI and sets build name and number from the job, so jf commands in the pipeline record build info automatically.

Q48. How do you integrate with GitHub Actions?

Use JFrog's setup action to install and configure the CLI, ideally with OpenID Connect so the workflow gets a short lived token instead of a stored secret.

Q49. What happens to JFrog Pipelines jobs today?

JFrog Pipelines had a feature freeze on 1 November 2024 and reached end of life on 1 May 2026, with no further fixes. Migrate the jobs to another CI tool such as GitHub Actions, GitLab CI or Jenkins and keep using the JFrog CLI for artifacts.

Q50. What happened to Mission Control?

From Artifactory 7.27.3, Mission Control runs as a microservice inside Artifactory, so it no longer needs a separate install. Some Mission Control user plugins were removed in later Artifactory versions.

Q51. How do you search for artifacts with the CLI?

jf rt search "libs-release-local/com/acme/*.jar" --props "build.name=api", alias jf rt s.

Q52. What is the role of the Access service?

It manages users, groups, permissions and tokens for the whole platform, so Xray, Distribution and Artifactory share one identity and permission model.

Q53. What are projects in the JFrog Platform?

A way to group repositories, builds, members and roles under one team or product, with project admins who manage them without global admin rights.

Q54. How do you back up Artifactory?

For small instances, the built-in system and repository backups. For large ones, back up the database and the filestore with your storage and database tools at a consistent point, because full exports become too slow.

Senior (5+ years)

Q55. Describe Artifactory's architecture.

A set of microservices (Router, Access, Artifactory, Metadata, Event and others) in front of a relational database for metadata and a filestore for binaries. The filestore can be local disk, NFS or object storage such as S3, Azure Blob or Google Cloud Storage.

Q56. Which database would you use in production?

An external PostgreSQL (or another supported database) with backups and monitoring. The bundled Derby database is for evaluation only.

Q57. How does Artifactory high availability work?

Several nodes share the same database and filestore behind a load balancer. All nodes serve traffic, and losing one node does not stop the service.

Q58. How do you size the filestore for growth?

Use object storage with a local cache on each node for hot binaries. Watch cache hit ratio, request latency and storage growth per repository, and enforce retention so snapshot repositories do not grow without limit.

Q59. How do federated repositories resolve write conflicts?

If the same file is deployed at two sites at once, the last event received wins, based on timestamps. Clock synchronization across sites is therefore critical.

Q60. Federation or replication?

Federation is bidirectional and keeps all members equal. Push and pull replication are one directional and suit a central source feeding read sites. Choose federation when teams in several regions write to the same repository.

Q61. How do you design repository naming?

Use a pattern such as team-tech-maturity-locator, for example payments-docker-prod-local. It makes permissions, cleanup and Xray watches easy to target with patterns.

Q62. How do you enforce immutable releases?

Give CI deploy rights only to staging repositories, promote to release repositories with a service account, deny overwrite and delete there, and use signed release bundles for what goes to production.

Q63. How would you block a newly disclosed critical CVE across the company?

Make sure all relevant repositories are in an Xray watch with a security policy that blocks downloads at that severity, use Xray to find every artifact and build that contains the component, and add the package to Curation policies so it cannot be cached again.

Q64. How do you produce an SBOM for a release?

Xray can export component details for artifacts and builds in SBOM formats such as CycloneDX and SPDX. Attach the SBOM to the release bundle or publish it with the build.

Q65. How do you keep pipelines working if Docker Hub limits pulls?

Route all pulls through a Docker remote repository behind a virtual repository. Cached layers are served locally, and only new layers go upstream with an authenticated account.

Q66. How do you monitor Artifactory?

Collect the platform's metrics and logs into your monitoring stack, and alert on request latency, error rates, database connection pool usage, filestore free space, garbage collection duration and replication lag.

Q67. Artifactory is slow during peak builds. Where do you look first?

Request logs for slow endpoints, database latency and connection pool exhaustion, filestore and cache latency, and any large AQL or search queries from scripts. Then check JVM memory and thread usage on each node.

Q68. How do you upgrade with minimal downtime?

Read the release notes for breaking changes, test in staging with a copy of production config, then upgrade HA nodes one at a time behind the load balancer. Back up the database before starting.

Q69. How do you secure the platform?

SSO with MFA, groups instead of users, scoped tokens with expiry, TLS everywhere, anonymous access off, admin actions audited, Xray and Curation enforced, and the database and filestore encrypted at rest.

Q70. How do you run Artifactory on Kubernetes?

Use the official JFrog Helm charts with an external database and object storage. Check the platform's minimum Kubernetes and Helm versions for your release, and give pods stable network addressing as the network requirements describe.

Scenario questions

Q71. A build passes today but fails tomorrow with the same code. What do you check?

Compare the two builds' build info. A different resolved dependency version, usually from a version range or a changed latest tag, is the most common cause. Pin versions and resolve through a virtual repository you control.

Q72. A developer overwrote a release artifact. How do you prevent it?

Remove overwrite and delete permissions from release repositories for everyone except a break glass admin group, and deploy releases only through promotion.

Q73. Disk is full but the team deleted thousands of files yesterday.

Empty or shorten the trash can retention, then run garbage collection. Check whether the deleted binaries are still referenced from other repositories through copies.

Q74. npm installs fail because the upstream registry is down.

If packages are already cached in the remote repository, serve them by enabling offline mode or adjusting metadata retrieval cache periods. Long term, keep all installs going through the virtual repository so the cache is warm.

Q75. Security asks which production services include a vulnerable logging library.

Use Xray's component search to find every artifact and build containing the vulnerable versions, then map builds to services through build info and release bundles.

Q76. Two regions need the same Docker images with low latency.

Use federated repositories if both regions publish, or Distribution to edge nodes if one region publishes and the other only consumes.

Q77. A token used by CI was leaked in a log.

Revoke the token, issue a new scoped token with expiry, check the access logs for its use, and move CI to OIDC based short lived tokens so there is no stored secret to leak.

Q78. Xray blocks a download the team says is a false positive.

Review the CVE's applicability, and if the risk is accepted, create an ignore rule scoped to that component, version and watch, with an expiry and a documented reason, rather than disabling the policy.

Q79. Migrating from Nexus to Artifactory: how do you plan it?

Inventory repositories and package types, map them to local, remote and virtual repositories, migrate hosted artifacts and their metadata with JFrog's migration tooling or the CLI, switch clients to virtual URLs, and run both systems in parallel until builds are green. See our Artifactory vs Nexus comparison.

Q80. Pipelines on JFrog Pipelines must move before support issues bite.

List each pipeline's steps, rebuild them in the target CI tool, keep artifact and build info steps on the JFrog CLI, run both in parallel for a release cycle, then retire the old pipelines.

Q81. Builds randomly fail with 401 errors after an SSO change.

CI jobs were probably using user credentials tied to the old identity source. Move them to service tokens or OIDC, and check group mappings so permissions still resolve.

Q82. Leadership wants to know how much storage each team uses.

Use storage summary data per repository, and if repositories follow a team naming pattern or projects, aggregate by team. Then set retention targets per team.

How to prepare

  • Install a trial or a local Artifactory and create one local, one remote and one virtual repository for Maven and Docker.
  • Publish build info from a sample project and promote it.
  • Configure an Xray watch and policy and trigger a violation with a known vulnerable dependency.
  • Read our container registry security practices for the registry side of these questions.

Artifact management with Artifactory is part of our DevOps training, in Kharadi, Pune and live online across India.

Frequently asked questions

How many JFrog questions are on this page?

82: 24 for freshers, 30 for 2 to 5 years, 16 for senior roles and 12 scenarios.

Which JFrog topics are asked most in interviews?

Repository types and virtual resolution order, build info and promotion, permissions and tokens, Xray watches and policies, and cleanup with garbage collection.

Is JFrog Pipelines still used?

JFrog Pipelines reached end of life on 1 May 2026 and no longer receives updates or security patches. Teams are migrating to other CI tools and keep the JFrog CLI for artifact steps.

What is the difference between Xray and Curation?

Curation blocks risky open source packages at the remote repository before they enter your organization. Xray scans artifacts and builds already stored and enforces policies on them.

What is a virtual repository in Artifactory?

A single URL that aggregates local and remote repositories. It resolves from local repositories first, then remote caches, then the remote sources.

Should CI use API keys for Artifactory?

No. JFrog has deprecated API keys. Use scoped access tokens with an expiry, or OIDC based short lived tokens where your CI supports it.

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.