Hello,
We use Jira Cloud and want to integrate it with our self-managed GitLab instance, which is hosted on our internal network behind a firewall. Our security team has not approved the integration yet because it may require exposing GitLab to inbound traffic from the internet. We would appreciate guidance on the following:
- Inbound access requirements
According to the GitLab documentation, the GitLab for Jira Cloud app requires the self-managed instance to be publicly reachable and to allow both inbound and outbound connections with Jira. Is this still required today? Which Jira Cloud features (e.g., the Development panel showing branches, commits, and merge requests) depend on Jira Cloud initiating connections to our GitLab instance? - IP allowlisting and endpoint restriction
If inbound access is unavoidable, can we restrict it to Atlassian's published IP ranges (and GitLab.com's ranges, if the app is installed through the GitLab.com proxy)? Which specific endpoints or paths on our GitLab instance must be reachable, so that we can block everything else (web UI, login page, git over HTTPS) at our reverse proxy or WAF? - OAuth scope and permissions
The setup uses a GitLab OAuth application with the "api" scope, which is broad. Is there a supported way to narrow the scope (e.g., read-only) or limit it to specific groups or projects? What data does Jira Cloud store from GitLab, and where is it stored? - Outbound-only alternatives
Is there an officially supported way to show GitLab development information in Jira Cloud using only outbound connections from our network? For example:
- GitLab's built-in Jira issue integration (API token, outbound only)
- Jira's OAuth credentials for on-premises tools, sending build, deployment, and development information from GitLab CI
Which features would we lose compared with the GitLab for Jira Cloud app
- Security documentation
Is there any official security documentation, architecture diagram, or data-flow description for this integration that we could share with our security team? - Recommended baseline security measures
Could you recommend the security measures that customers commonly put in place when they connect a self-managed GitLab instance to Jira Cloud? We are especially interested in recommendations for:- Network controls (firewall rules, reverse proxy or WAF configuration, IP allowlisting)
- Authentication and authorization (dedicated integration account, least-privilege permissions, token lifetime and rotation)
- Monitoring and auditing (which logs to review and how to detect unusual access from the integration)
- Patch management (how quickly to apply GitLab security releases while the instance is exposed)
- Any best-practice guide or reference configuration would help us a lot when we present this to our security team.
Environment:
- Jira Cloud (Software)
- GitLab self-managed
Thank you in advance for your help.