Apply Retention Policies Task never stops after Server Crash

Hello @ all,

Our System:

Octopus Deploy 2.5.12.666
Windows Server 2012 R2 Datacenter (64-bit)
12GB RAM
6x 2.7GHz Cores

Our Problem:

I reckognized that after a Server crash the “Apply Retention Policies” Task runs for several hours til we canceled this task manually. Last time that we’ve got this was today at 14:50 (UTC+1). This Problem is already known since we use version 2.5.8.447. It was already communicated (http://help.octopusdeploy.com/discussions/problems/24589-several-problems-with-octopus-deploy-258447).

Raw View of the Task that never stops

Task ID:       ServerTasks-45565
Task status:   Canceled
Task queued:   Tuesday, January 6, 2015 2:41 PM
Task started:  Tuesday, January 6, 2015 2:41 PM
Task duration: 1 hour
                    | Cancelled: Apply retention policies
14:41:59   Info     |   Loading retention policies...
14:41:59   Info     |   Applying retention policies...
14:41:59   Info     |   Note that this task only applies Octopus server-side retention policies for releases. Files on Tentacles will only be deleted when the retention policy is applied at the end of a deployment.
15:55:51   Info     |   Cancellation requested...
15:55:51   Info     |   Cancellation the task of is complete
                    | 
                    |   Cancelled: Apply policy Keep for 3 months
14:41:59   Info     |     Policy name: Keep for 3 months
                    |     Undeployed releases: Last 30 days
                    |     Deployed releases:   Last 90 days
                    |     Projects covered:
                    |     - 136 counted projects
14:41:59   Info     |     1470 Releases will be kept
                    |   

After I canceled this task I triggered it again and this time the task finished without any errors.

Raw View of the retriggered task

Task ID:       ServerTasks-45576
Task status:   Success
Task queued:   Tuesday, January 6, 2015 3:55 PM
Task started:  Tuesday, January 6, 2015 3:55 PM
Task duration: 19 seconds
                    | Success: Apply retention policies
15:55:55   Info     |   Loading retention policies...
15:55:55   Info     |   Applying retention policies...
15:55:55   Info     |   Note that this task only applies Octopus server-side retention policies for releases. Files on Tentacles will only be deleted when the retention policy is applied at the end of a deployment.
15:56:14   Verbose  |   All policies applied
                    | 
                    |   Success: Apply policy Keep for 3 months
15:55:55   Info     |     Policy name: Keep for 3 months
                    |     Undeployed releases: Last 30 days
                    |     Deployed releases:   Last 90 days
                    |     Projects covered:
                    |     - 136 counted projects
15:55:55   Info     |     1765 Releases will be kept
15:56:14   Info     |     21 releases were deleted
                    |   

The Crashdump was uploaded to “https://file.ac/cMkQMTlkzkY/” and the package name is “CID_CrashDump-OctopusServer-v2.5.12.666_01-06-2015_14-55.7z”.

If you need more informations please let me know.

Greetings Marco

2 new Dumps are uploaded… After both crashes the “Apply Retention Policies” Task getting stucked.

About the Server crash. We use every setting what was documented here: http://docs.octopusdeploy.com/display/OD/Switch+from+single-threaded+workstation+GC+to+concurrent+server+GC

Greetings Marco

Hi Marco

We’ve investigated the 2 new crash dumps that you have uploaded and to much dismay it is the same AccessViolationException that we have seen in the other crash dumps we have investigated as well. :frowning:

Unfortunately there isn’t much we can do about these crashes from occurring, but with 3.0 and our switch from RavenDB to SQL we’re confident that these types of crashes will be nevermore.

Sorry for the bad news.

Thank you and warm regards,

Henrik

Hi Henrik

Thanks for this information but you’re right. That’s really bad news. :frowning:
So I think we have to wait til 3.0 is released.

Greetings Marco