GitLab Upgrades

Upgrade GitLab.
Keep control.

Plan a self-managed GitLab upgrade around your installation. Agree the version path, checks and recovery procedures before the maintenance window.

Read the case study

Before the maintenance window

  1. Know the starting point

    Version, topology and dependencies

  2. Check readiness

    Backups, restore and stop/go criteria

  3. Validate the change

    Instance, pipelines and integrations

Illustrative checkpoints. The actual route depends on your installation.

A customer story

Version 14.
A complete upgrade journey.

Financial services organisation

A financial services organisation was running GitLab on-premises at version 14. Ageing hardware, slow builds and unexpected downtime were affecting developers. Security patches had not been applied, newer features were out of reach, and an undetected HTTP vulnerability remained open in the network.

OTTRA upgraded the instance to the then-current GitLab release, restoring security patch coverage and unlocking AI features without increasing licensing costs. We established a rolling 12-month programme of zero-touch upgrades through a dedicated runner connected to the customer’s infrastructure. The customer controlled when upgrades ran; hypercare support and immediate rollback capability were part of that programme.

During the engagement, OTTRA identified and resolved the HTTP vulnerability. We also recommended moving from a High Availability configuration to a Geo instance on AWS reference architecture, significantly reducing monthly infrastructure costs while improving reliability.

A planned change.
A checked result.

Several versions behind, facing required upgrade stops or unsure what depends on the instance? Start with its version and topology, using current GitLab guidance.

An installation-specific plan

Record the current version, deployment method, topology and dependencies. Identify the supported target, required stops, prerequisites and acceptance checks.

An executed, validated upgrade

Run the agreed sequence in the scheduled window. Check the instance, representative pipelines and important integrations against the plan.

A usable handover record

Document the result, known issues, remaining actions and owners. Ongoing maintenance has its own schedule, responsibilities and support boundaries.

How we prepare, run and hand over the change

Agree access, backup and restore procedures, stop/go criteria and who can authorise recovery. Rehearsal needs depend on the environment.

Set the maintenance window, communications, validation and support coverage with the platform owners. Record the checks and hand over remaining actions. Further maintenance is a separate choice.

Which version are
you working from?

Bring your GitLab version, installation method, topology, key integrations and available maintenance windows.

Discuss your upgrade

We agree scope, timing and cost before work starts.