Uploaded image for project: 'Bamboo Data Center'
  1. Bamboo Data Center
  2. BAM-17847

Dont remove logs from failed deployments if is included by Minimum to keep

      Summary

      Failed logs can be extremely important as the logs for a successful deployment. Bamboo seems to expire these currently even when a deployment falls under "Minimum Deployments To Keep".

      Environment

      Bamboo 5.12.X
      Bamboo 5.9.X

      Steps to Reproduce

      1. Set the global expiration settings to:
      Expire After: 0 days
      Min builds to keep: 3
      Min deployments to keep: 3
      2. Run a few deployments, then for the last 3 have one fail and 1 sucessful.
      3. Manually run Expiraiton.

      Expected Results

      Most customers would expect that all logs should be preserved as the deployments have been preserved.

      Actual Results

      Deployments beyond the minimal 3 are correctly deleted. However, the logs for the failed builds are cleared - but the successful build is kept.

      Workaround

      Not ideal, but extend the value for Expire After the logs for longer.

            [BAM-17847] Dont remove logs from failed deployments if is included by Minimum to keep

            Marcin Gardias added a comment - - edited

            Hi,

            Unfortunately, this is not a straightforward fix. Our intent was to preserve n latest successful deployments. We cannot change the meaning of the option now to n latest any deployments, as this would break the existing behavior for those that rely on it.

            IMO the solution is to introduce a second number, so the expiry would be able to keep n successful and m of any results.

            We are going to add the feature to a 7.0.x point release. Default value for the new setting will be '0' (in order to preserve the current behavior), so after installing the fix, you'll need to set the value manually.

             

            Marcin Gardias added a comment - - edited Hi, Unfortunately, this is not a straightforward fix. Our intent was to preserve n latest successful deployments. We cannot change the meaning of the option now to n latest any deployments, as this would break the existing behavior for those that rely on it. IMO the solution is to introduce a second number, so the expiry would be able to keep n successful and m of any results. We are going to add the feature to a 7.0.x point release. Default value for the new setting will be '0' (in order to preserve the current behavior), so after installing the fix, you'll need to set the value manually.  

            Our teams struggling on finding the root cause for failures as logs are getting cleared and no other option(s) to know the cause for failures. Hope this issue addressed at the earliest.

            Kiran Srikakolupu added a comment - Our teams struggling on finding the root cause for failures as logs are getting cleared and no other option(s) to know the cause for failures. Hope this issue addressed at the earliest.

            This issue is still exists in 6.7.3+.

            Anuroop Kottamparambil added a comment - This issue is still exists in 6.7.3+.

              mgardias Marcin Gardias
              nhickey Neil Hickey
              Affected customers:
              12 This affects my team
              Watchers:
              10 Start watching this issue

                Created:
                Updated:
                Resolved: