Skip to content
LogluneStart free

Blog

Reproducing the bugs that only live in one customer’s configuration

Guides for reproducing customer-specific bugs, responding to incidents, releasing changes safely, and running branches and pipelines, plus comparisons of feature-flag platforms on pricing, hosting, and past-state reproduction.

01

Reproduce bugs and respond to incidents

Guides for rebuilding the flag state a failing request evaluated, running incident response, and measuring time to recovery.

10 articles

  1. Application Performance Monitoring: How APM Works

    For teams that run web or backend services: learn what APM measures, how it works, how to choose a tool, and how to trace a regression to a flag change.

  2. Bitemporal Data: Correct the Past, Keep the Record

    Decide whether a table needs two time axes. Learn how bitemporal data corrects a past value without losing what you knew, and how to store it.

  3. Event Sourcing: Store Events, Rebuild Past State

    Event sourcing stores each change as an immutable event and rebuilds state by replay. Decide if it fits, with a real run that rebuilds a flag at time T.

  4. git checkout: Switch Branches and Restore Files

    Use git checkout to switch or create branches, check out a remote branch, restore a file, or inspect an old commit, and see when git switch fits better.

  5. Slowly Changing Dimension: Choose and Build Type 2

    Engineers choosing an SCD type per attribute can use a decision table for Types 0 to 7, then build Type 2 in SQL, Delta Lake, or dbt and query it by date.

  6. Temporal Database: Query Data as of Any Past Time

    For developers choosing how to keep history and query it as of a past time T: valid and transaction time, SQL:2011 tables in four databases, AS OF queries.

  7. git bisect: Find the Commit That Introduced a Bug

    Find the commit that introduced a bug with git bisect: run 6 steps by hand, automate them with git bisect run, and bisect flag edits when no commit is bad.

  8. Incident Management: Set Up a Process in 6 Steps

    For software teams with an on-call rotation: set up incident management in 6 steps, from severity levels and roles to restoring service and the review.

  9. MTTR: How to Measure and Reduce It in 7 Steps

    For teams that run software services: pick which MTTR you report, record the right timestamps, calculate the mean, and cut the stage that takes longest.

  10. Reproduce a Bug That Only Happens for One Customer

    For developers whose bug fires in one customer's account only: rebuild the customer's flag state at the moment it failed and replay it in dev, CI, or staging.

02

Release changes safely

Guides for separating deploy from release with feature flags, canary and blue-green deployment, and beta programs.

12 articles

  1. Changelog: What to Record and How to Write One

    For maintainers who keep a CHANGELOG.md: the format, the 6 change categories, an example, 6 writing steps, and how to log changes shipped behind flags.

  2. Use git commit --amend to Fix Your Last Commit

    Use git commit --amend to fix the last commit's message, files, or author, see when amending a pushed commit is safe, and undo an amend from the reflog.

  3. git rebase: Update a Branch and Clean Up Commits

    Use git rebase to update a branch from main, resolve conflicts, squash or drop commits with interactive rebase, undo a rebase, and know when to merge instead.

  4. git reflog: Find and Recover Lost Commits

    Use git reflog to find commits lost after a hard reset, a rebase, an amend, or a deleted branch, read names like HEAD@{2}, and go back without losing work.

  5. git reset: Go Back a Commit and Undo a Reset

    Use git reset to go back to an earlier commit with --soft, --mixed, or --hard, unstage a file, match origin, and undo a reset that went wrong with git reflog.

  6. git revert: Undo a Commit and Keep the History

    Use git revert to undo a pushed commit with a new commit: revert the last commit, a range, or a merge with -m, resolve conflicts, and choose revert or reset.

  7. git tag: Create, Push, and Check Out a Release Tag

    For developers who mark releases in Git: create an annotated tag, push it, check it out, and fix a release with a new tag instead of moving the old one.

  8. How to Write Release Notes: Template and Example

    For product teams writing release notes: what to include, a copyable template and example, 6 writing steps, and how to list changes shipped behind flags.

  9. Beta Testing: 7 Steps From First Tester to Release

    Beta testing gives a pre-release version to a limited group of users. App and web teams follow 7 steps to pick testers, collect feedback, and end the beta.

  10. Blue-Green Deployment: 7 Steps to a Quick Cut-Over

    A blue-green deployment switches all traffic to a second, identical environment. Web teams can follow 7 steps to build green, switch, and roll back.

  11. Canary Deployment: 6 Steps to Roll Out in Stages

    A canary deployment tests a change on a small part of production. Teams that run their own service can follow 6 steps to check it and roll forward or back.

  12. Feature Flags: 7 Steps From First Flag to Removal

    Feature flags switch which code path a user gets without a deploy. Follow 7 steps from flag type and targeting to rollout, past-state checks, and removal.

03

Branching, pipelines, and delivery metrics

Guides for choosing a branching model, building a CI/CD pipeline, and measuring delivery with the DORA metrics.

12 articles

  1. Audit Logging: What to Record and How to Secure It

    For teams that must show who changed what: decide what an audit log records, make it tamper-evident with append-only rows and a hash chain, and set retention.

  2. CQRS: When to Split Your Read and Write Models

    CQRS splits writes into a command model and reads into a query model. Decide whether the split is worth its added complexity and eventual consistency.

  3. git diff: Compare Commits, Branches, and Files

    Use git diff to check changes before a commit, compare two commits, branches, or files, read the unified output, and fix the ^M carriage returns in a diff.

  4. git log: View and Filter Your Commit History

    Use git log to read the commit history, show the last 10 commits, draw branches with --graph, and filter commits by date, author, message, code, or file.

  5. Automated Tests: How to Write and Run Them in CI

    For teams that ship web services from one mainline: write an automated test, run the suite in CI from fast to slow, and fix the inputs that make tests flaky.

  6. Regression Testing: How to Run It in 5 Steps

    For teams that ship web services and APIs through CI: learn what regression testing covers, pick a type, fix the inputs, and run a suite on every change.

  7. Testing Flags: How to Test Feature-Flagged Code

    Testing flags means testing code behind a feature flag. Unit test each flag value, then pin the flag version production served in integration tests.

  8. Unit Tests: How to Write and Run Them in 5 Steps

    For developers who want tests that give the same result on every run: write unit tests with Arrange, Act, Assert, pass hidden inputs in, and run them in CI.

  9. DORA Metrics: Measure and Improve Them in 7 Steps

    For teams that ship one service: measure the five DORA metrics with fixed counting rules, decide how flag changes count, and improve one constraint at a time.

  10. CI/CD Pipeline: How to Build One in GitHub Actions

    A CI/CD pipeline builds, tests, and deploys each change. Learn CI vs. CD, then build a GitHub Actions pipeline that tests pull requests and ships one build.

  11. Gitflow: 7 Steps to Branch, Release, and Hotfix

    Gitflow is Vincent Driessen's 2010 branching model with master, develop, and supporting branches. Teams shipping versioned software get 7 steps to use it.

  12. Trunk-Based Development: 7 Steps to Merge Daily

    Trunk-based development means merging small changes into one shared trunk every day. Teams on long-lived branches can follow 7 steps to make the switch.

04

Choose a feature-flag platform

Platforms compared on published pricing, hosting, and past-state reproduction.

1 article

  1. 8 LaunchDarkly Alternatives in 2026 by Pricing

    For teams leaving LaunchDarkly: 8 alternatives compared on published price, self-hosting, and whether each can rebuild a customer's past flag state.

Reproduce a customer’s flag state

Pick a customer and a time or version. Loglune evaluates the flag state that was live then, and your SDK gets those answers in dev, CI, or staging.

Start freeSee the console