Details
-
Bug
-
Resolution: Unresolved
-
P1: Critical
-
None
-
master
-
None
-
macOS
Description
Sometimes the provisioning log of a provisioning work item will say that everything is fine, but in the Coin UI the work item is marked as Error.
http://coin/coin/integration/qt/qtbase/tasks/web_qt_qtbase_1726049990853
The log ends with:
agent:2024/09/11 08:53:30 agent.go:387: Provision succeeded INFO:workitem.py:691(100043): Agent FINISHED PASS: provisioning/qtci-macos-15-arm-103-95a4b6/Provision INFO:workitem.py:773(100043): <a href="https://inframetrics.intra.qt.io/d/nOAsINmZz/telegraf-coin-vms?orgId=1&from=1726043635000&to=1726044831000&var-server=1725950057-25405">VM metrics</a> INFO:workitem.py:776(100043): <a href="https://inframetrics.intra.qt.io/d/nOAsINNZz/telegraf-hosts?orgId=1&from=1726043635000&to=1726044831000&var-server=qt-srv-arm-mac29">Host metrics</a> INFO:workitem.py:975(100043): Booting up template 'qtci-macos-15-arm-103-95a4b6' (agent: 1725950057-26441)...
Yet
The working theory is that this only happens with the Tart integration, due to possibly:
- New Tier2 push to registry fails. Newer versions of Tart seems to have improvements for that.
- New Tier2 created and heavy load in CI leads Tier2 pull from registry to take too long and COIN timeouts.
The first step to solve this is to add logging at the places that does the Tier2 push and pull, so we can verify that this is what happens.
Attachments
Issue Links
- depends on
-
QTQAINFRA-6593 Update Tart hosts and Tart itself to newest versions
-
- Closed
-