Project

General

Profile

Actions
Copy link

action #36565

closed

parallel parents aren't killed when chained parents fail

Added by coolo almost 6 years ago. Updated over 5 years ago.

Status:
Resolved
Priority:
High
Assignee:
coolo
Category:
Regressions/Crashes
Target version:
Done
Start date:
2018-05-27
Due date:
% Done:

0%

Estimated time:

Description

we have ha-install (https://openqa.suse.de/tests/1719787) that failed which
lead to the chained children ha-node1 and ha-node2 to be cancelled. But
the parallel parent to them ha-supportserver is left scheduled (https://openqa.suse.de/tests/1719788).

Just as with restart, we need to cancel the full cluster if jobs fail.

Actions
Copy link
 #1

Updated by coolo almost 6 years ago

  • Assignee set to coolo
  • Target version changed from Ready to Current Sprint
Actions
Copy link
 #2

Updated by coolo almost 6 years ago

  • Status changed from New to Resolved
Actions
Copy link
 #3

Updated by szarate almost 6 years ago

  • Target version changed from Current Sprint to Done
Actions
Copy link
 #4

Updated by AdamWill over 5 years ago

I think the fix for this went a bit too far, and causes the behaviour I just filed as an issue itself:

https://progress.opensuse.org/issues/46295

it seems that this code means that if you just have three jobs that are parallel to each other, and any one of them fails, the other two are immediately cancelled as parallel_failed . But that does not necessarily make sense at all.

Actions
Copy link

Also available in: Atom PDF