Running on Bitbucket?

Bitbucket Pipelines or runners misbehaving? We unstick it.

Bitbucket often arrives in the same breath as a Jira project and a trail of half-documented repositories. The pipelines were written by someone who left, the self-hosted runners drift offline for reasons nobody logs, and the LFS quota has been maxed twice this year. We tidy the pipelines, keep the runners up, and bring the whole git side of the Atlassian stack back under control.

  • Since 2015Fixing technology
  • FreeEstimates
  • US teamReal engineers

bitbucket-pipeline.log

still failing

What it is

Bitbucket at a glance

Bitbucket is Atlassian's hosted git service, aimed at teams that already use Jira and Confluence. It launched in 2008 as a Mercurial host and was acquired by Atlassian in 2010. Today it offers git repositories, pull requests, branch permissions, and CI through Bitbucket Pipelines, with cloud and self-hosted Data Center editions.

Pipelines are declared in a bitbucket-pipelines.yml at the repo root. Steps run in Docker images on Atlassian-hosted Linux runners, on your own self-hosted Linux, macOS or Windows runners, or on a mix of both. The deep hooks into Jira, including branch creation from issues and automatic transitions, are the main reason teams on the Atlassian stack stay with it.

Type
Hosted git repository service with integrated CI
First released
2008, acquired by Atlassian in 2010
License
Proprietary SaaS, Data Center for on-prem
Maintained by
Atlassian
Typically used for
Git hosting, pull requests, pipelines, deployments for teams on the Atlassian stack

The honest version

Bitbucket pros and cons

No technology is all upside. This is where Bitbucket earns its place and where it tends to cause trouble.

Where Bitbucket is strong

  • Deep Jira integrationBranches can be created from a Jira ticket, pull requests update the ticket state, and deployments close issues automatically. Teams that live in Jira get real value out of this loop.
  • Free private repos for small teamsThe cloud free tier still includes unlimited private repositories for up to five users, which is a straightforward win for small shops or side projects.
  • Branch permissions are granularWrite access can be scoped per branch or branch pattern, which means production branches can be locked while feature work remains open.
  • Self-hosted runners work wellLinux, macOS and Windows runners can live inside your network, pulling jobs from the cloud scheduler without opening inbound ports.
  • Deployments as first-class objectsEnvironments, approvals and deployment history live in the UI, not just in a YAML file, which makes release conversations with non-engineers easier.
  • Data Center edition for regulated shopsTeams who cannot use cloud get a self-hosted edition that still looks like Bitbucket Cloud, with the same workflows and roughly the same UI.

Where Bitbucket bites

  • Pipeline minutes run out faster than people expectThe default plan ships with a modest build minutes allowance. A monorepo with any real test suite will eat through it in a week, and the top-up pricing is not cheap.
  • Ecosystem is smaller than GitHubFewer third-party integrations, fewer Marketplace apps, and more hand-rolled glue code when you want a tool that assumes GitHub.
  • Pipelines YAML is less flexible than competitorsReusable steps, matrices and conditional logic are more limited than GitHub Actions or GitLab CI. Complex workflows end up duplicating config.
  • LFS quota and storage costs surprise large reposGame teams, design-heavy repos, and anyone with binary assets will hit LFS limits and pay a storage uplift that is harder to predict than the minutes bill.
  • Smart Mirrors need care in multi-region setupsData Center mirrors drift when the primary is under load, and the recovery steps are not something you want to meet for the first time during an incident.
  • UI changes bite workflows that depended on themSeveral redesigns in recent years have broken bookmarks, deep links from Jira, and third-party browser extensions. Teams do not love surprise churn.

Sound familiar?

Common Bitbucket problems we fix

The Bitbucket problems that bring people to us, and what you have once each one is fixed.

Could not pull Docker image

Pipeline cannot start the step

A private registry credential missing from the pipeline variables, a base image that was moved, or Docker Hub rate-limiting an anonymous pull. The step dies before any user code runs.

Images pulled, builds running

Minutes used 2,480 of 2,500

Build minutes exhausted mid-month

A matrix of tests running on every push, no path filters on feature branches, and every PR running the full E2E suite. The pipeline grinds to a halt until the next billing cycle.

Minutes trimmed, releases unblocked

Runner 'linux-self-hosted-1' is offline

Self-hosted runners go silent

A systemd service that stopped on reboot, a token that rotated without the runner picking it up, or a network path to the cloud that drops at random intervals.

Runners healthy and self-healing

LFS storage quota exceeded

Large binary assets block the push

A design folder pushed without LFS pointers, or an LFS history that nobody has pruned since the project started. The workspace quota locks writes until it is cleared.

LFS tidied, quota under control

Pipeline variable not defined

Secrets are missing or scoped wrong

A deployment variable set on one environment but referenced in another, or a repository variable that was deleted during an audit. Deploys fail without a helpful message.

Secrets scoped per environment

Service 'postgres' failed to start

Service containers refuse to boot

A Postgres or Redis service container running out of memory, misconfigured, or on a version that no longer matches the application. The step fails after the service health check times out.

Services stable through the step

Step exceeded 120 minute limit

Builds time out at the hard ceiling

A test suite that has grown past the cloud runner cap, a flaky end-to-end step that retries forever, or a Gradle build with no daemon caching between runs.

Builds finish inside the limit

Branch permissions prevent push to 'main'

Releases are blocked by branch rules

A protected branch set to require reviewers or Jira linkage that the deployment user cannot satisfy. The pipeline passes CI and then fails to push its own version bump.

Rules that match the real workflow

What we do

Bitbucket help, from one broken pipeline to the whole workspace

Hand us a single failing pipeline, or let us look after the whole workspace.

Rescue
Pipelines red, runners offline, or LFS quota blocking pushes. We get the team unblocked within hours and tidy properly afterwards.
Pipeline rewrites
A sensible bitbucket-pipelines.yml with path filters, reusable step anchors, and matrix builds only where they earn their keep.
Self-hosted runners
Linux, macOS and Windows runners deployed with autoscaling, health checks, and a watchdog that restarts them before an engineer notices.
Deployments and approvals
Environments, gates and approvals configured against your real release process, with Jira transitions that match.
Migrations
Moving into Bitbucket from GitHub or GitLab, or away from Bitbucket when the fit is no longer right, with history preserved.
Managed Tech
Pipeline health watched, minutes usage reviewed weekly, runners kept patched, and a human on call when a deploy decides Monday is a good day to fail.

How it works

From first message to fixed

  1. Look Free estimate

    Tell us what's going wrong. We look at it, then tell you plainly what's wrong and what it will cost.

  2. Fix

    We fix the cause, not the symptom. If something is down, getting it back up comes first.

  3. Harden

    Backups, monitoring, updates and access, so the same problem can't quietly come back.

  4. Hand back or stay on

    You get a plain-language write-up. Or we stay on for hosting, upkeep and new work.

Often found together

Running Bitbucket alongside something else?

We work on the whole stack, so the problem can be wherever it actually is.

Questions

Bitbucket questions

Something else? Call+1 737 727 7327

Why do my Bitbucket Pipelines use so many minutes?

Three common causes: no path filters so every push triggers the full workflow, matrix builds that multiply steps by OS and language version, and e2e tests running against every branch rather than only release candidates. We audit the recent month, point at the specific pipelines doing the spending, and trim without slowing the release train.

Can I use self-hosted runners with Bitbucket Cloud?

Yes. Linux, macOS and Windows runners can register with Bitbucket Cloud and pull jobs over an outbound connection, so no inbound firewall opening is needed. Teams run them on their own VMs, bare metal, or in autoscaling groups when the load is bursty.

How do I migrate from GitHub Actions to Bitbucket Pipelines?

The YAML syntax is different but the concepts overlap: steps, caches, services, deployments and approvals all have equivalents. The sharp edges are usually marketplace actions that have no Bitbucket twin and matrix workflows that need to be flattened. We run the new pipeline in parallel for a release cycle before switching away from GitHub.

Why did our LFS quota suddenly fill up?

Usually a binary asset committed without the LFS pointer in place, or a long history of design files that was never pruned. Rewriting the history to move those files into LFS, and then running a garbage collection on the server side, usually recovers the space, but it needs a planned window because it rewrites commit hashes.

Does Bitbucket support signed commits and required reviewers properly?

Yes, verified author emails, signed commits and branch permissions can be enforced on protected branches. The caveat is that default branches created before the policy change are not retroactively locked, so new rules have to be applied with a review of which branches already exist.

How much does Bitbucket help cost?

We quote at no charge. Send over the workspace name, a pipeline URL that is failing, and a short note on what you want fixed, and a price comes back the same day.

Free estimate

Bitbucket giving you trouble?

Tell us what you're running and what's going wrong. A person reads it and comes back with an estimate.

  • No charge to find out what it will take
  • If it's urgent, say so

Replies come from a person. By sending, you authorize us to email you back. No marketing.