juju 2.2.4 and 2.2.6 actions dissappear when state is changed from running to complete
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Canonical Juju |
Fix Released
|
High
|
Eric Claude Jones |
Bug Description
Description: After 20ish hours of calling actions in a while loop that take either a minute, 30 minutes, an hour, or 4 hours to run. The result of an action disappears and juju loop.run gets stuck in an infinite loop.
Code:
def run_action(self, unit, action, params):
from juju import loop
app = self.model.
action = loop.run(
return action
while True:
Log from above code with timestamp of the start of the action:
jujubench2.
$ juju debug-log -m sshdefault
unit-sshdefault-0: 17:46:47 ERROR juju.worker.uniter resolver loop error: executing operation "run action b57d6c41-
unit-sshdefault-0: 17:46:47 ERROR juju.worker.
$ juju show-action-status -m sshdefault b57d6c41-
ERROR no actions found matching prefix "b57d6c41-
Note: The errors show up in the juju debug-log after the action completes, but I can not find the action using juju show-action-status. This indicates to me that when the action is moved from the running to completed state it gets lost.
Changed in juju: | |
milestone: | 2.3-beta3 → none |
Changed in juju: | |
status: | Triaged → In Progress |
Changed in juju: | |
status: | In Progress → Fix Committed |
Changed in juju: | |
milestone: | none → 2.3.5 |
tags: | added: sts |
Changed in juju: | |
status: | Fix Committed → Fix Released |
I have tried this with both a default bootstrap of the juju controller, as well as a bootstrap of a controller with the following config values.
automatically- retry-hooks: false os-refresh- update: false vendor- metrics: false
enable-
enable-os-upgrade: false
firewall-mode: none
test-mode: true
transmit-