Issues
Issues copied to clipboard
OctopusDeploy
Unresolved manual intervention holding transaction open forever
Team
- [X] I've assigned a team label to this issue
Severity
kinda non-blocking
Version
Reported on 2021.2.7650
Latest Version
No response
What happened?
Unattended manual interventions after some time cause db (replicated) to time out. The tempdb version stores grow too large. We believe there's a transaction somewhere being kept open while waiting for the interruption to resolve.
Reproduction
Below is a best guess
- Create a project with a simple deployment process that would error
- Use guided failure
- Deploy, which should result in a manual intervention
- Wait for a long time and watch the tempdb version stores grow in size
Error and Stacktrace
No response
More Information
Internal slack discussion - https://octopusdeploy.slack.com/archives/CNHBHV2BX/p1656639072790969 More discussion - https://octopusdeploy.slack.com/archives/CNHBHV2BX/p1660868210247089 Customer report - https://octopus.zendesk.com/agent/tickets/92111
Workaround
Customer suggests restarting the mssql services resolves the issue at the cost of losing the intervention task, so is not a full workaround.
Another affected customer: https://octopus.zendesk.com/agent/tickets/92598
Another customer is facing this error: SQL Error 1222 - Lock request time out period exceeded
Their issue could be related to this manual intervention problem: [internal] ticket.
Release Note: Fixed an issue where a long-running database transaction would grow the transaction log unreasonably large during a long deployment
:tada: The fix for this issue has been released in:
| Release stream | Release |
|---|---|
| 2022.3 | 2022.3.11320 |
| 2022.4 | 2022.4.8583 |
| 2023.1 | 2023.1.9794 |
| 2023.2 | 2023.2.5159 |
| 2023.3+ | all releases |
