Auto-Deploy unreliable in elastic tenanted environment

execution

(wilson85) #1

We have a tenanted elastic environment set up with auto triggers for when ‘machine becomes available for deployment’ and we’re noticing issues when we boot multiple new servers up, some servers often do not get deployed to at all. On inspecting the auto-deploy logs, i can see this
The machine abc is not eligible for the auto-deployment of release Releases-31651 (Projects-309, Environments-43, Tenants-1) because a deployment is currently in progress. abc will receive the deployment when it completes.

This machine however never gets deployed to.

If i look at deployments to that specific server, i can see that nothing was in progress when it reported above. The issue always seems to be with the last server to get booted, which makes me think there is a bug when a server becomes available mid ‘processing automatic deployments for triggers’.

We recently upgraded from 2018.6.1 where we never saw this issue.


(Shane Gill) #2

Hi @wilson85,

Thanks for getting in touch.

I have been trying to reproduce this issue and can’t seem to get it.

Could I please get your auto deploy logs when this happens? They are typically located in C:\Octopus\TaskLogs and begin with scheduledtasks_processautodeployments. You can send them to support@octopus.com to protect your privacy, please reference this issue when sending them.

Also, which version have you upgraded to?

Thanks,
Shane


(wilson85) #4

I’ve sent the log.

The problem entry seems to be at 2018-10-08T11:25:04.5898513+00:00


(Shane Gill) #5

Thanks for sending the log.

I have managed to reproduce and fix the issue: https://github.com/OctopusDeploy/Issues/issues/4973

The fix will be shipped shortly.

Cheers,
Shane


(Shane Gill) #6

The fix has been shipped in release 2018.8.12. Look forward to finding out if tenanted auto-deploy are more reliable now.

Cheers
Shane


(system) #7

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.