Block a user
Check for number of workers before soft failing the task.
@dr.sybren I have written the tests could you review it.
94848c7e05
Merge remote-tracking branch 'upstream/main' into 104190-job-stuck
0e0fddc076
Add-on: Change 'No Cluster' to 'All'
1da9c33f22
WorkersLeftToRun should return the UUID of the test worker which is actually failing the task in the test.
e0f1400f4d
Merge remote-tracking branch 'upstream/main' into 104190-job-stuck
b74db1daed
Clarify worker clusters in CHANGELOG.md
6a89fa346c
Manager: correctly count how many workers can run a job
b74db1daed
Clarify worker clusters in CHANGELOG.md
6a89fa346c
Manager: correctly count how many workers can run a job
5c101c47fb
Merge remote-tracking branch 'upstream/main' into 104190-job-stuck
a36c4cd4e4
Web: clarify the cluster assignment result in the worker details view
10d7e7e203
Manager: allow creation of worker clusters without UUID
f5ab2bb4c2
OAPI: regenerate code
0e6b6a6013
API: allow worker cluster creation without explicit UUID
a36c4cd4e4
Web: clarify the cluster assignment result in the worker details view
10d7e7e203
Manager: allow creation of worker clusters without UUID
f5ab2bb4c2
OAPI: regenerate code
0e6b6a6013
API: allow worker cluster creation without explicit UUID
3724a8874e
Slight change of worker cluster behaviour
ff0a36d19d
Add test to check the job failure condition when number of workers available for the job is less than failure threshold.
ac88d57ede
We should also hard fail the task when numFailed == threshold
ad96e3bb25
Merge remote-tracking branch 'upstream/main' into 104190-job-stuck
e2559b1181
Cleanup: remove doubly-declared default value in persistence layer
159ce5b34a
Manager: avoid starting error messages with 'error'
5ceafb1a9f
Merge remote-tracking branch 'upstream/main' into 104190-job-stuck
e2559b1181
Cleanup: remove doubly-declared default value in persistence layer
159ce5b34a
Manager: avoid starting error messages with 'error'
4a142316a1
Add
--tag
to git describe
996f5da2af
Bumped version to 3.3-alpha0
a386d80765
Bump govulncheck dependencies
1b60b43736
Update CHANGELOG
0ac64719e7
Job deleter: improve logging
Job can get stuck with less workers than soft-fail threshold
If number of workers capable of running the failed task again is "1" that means we have no worker besides the one that actually failed the task. In such condition we should just fail the job itself.
Check for number of workers before soft failing the task.
9fdf5aa7c5
Manager: fixed issue #104190 job getting stuck with less workers than soft-failed threshold