logsender gets wedged on shutdown sometimes
Bug #1732588 reported by
Tim Penhey
This bug affects 1 person
| Affects | Status | Importance | Assigned to | Milestone | |
|---|---|---|---|---|---|
| Canonical Juju |
Fix Released
|
High
|
Tim Penhey | ||
| 2.2 |
Fix Released
|
High
|
Tim Penhey | ||
Bug Description
Observed during prodstack 4.5 upgrade, a number of agents would get stuck attempting to restart for the agent upgrade.
The engine report and goroutine output from the agent pointed at the logsender.
It would get stuck initiating the websocket protocol where it was waiting forever for a response for the server after the server stopped listening to it, in a similar way to bug 1732587.
To see if this is what is causing the agent to be wedged, ssh in to the machine and run 'juju-goroutines'.
There should be approximately 15-20, and one will be in a tcp read from the logsender.
Revision history for this message
| Tim Penhey (thumper) wrote | #1 |
| Changed in juju: | |
| status: | In Progress → Fix Committed |
| description: | updated |
Revision history for this message
| Tim Penhey (thumper) wrote | #2 |
| Changed in juju: | |
| status: | Fix Committed → Fix Released |
To post a comment you must log in.
https:/