Let's create per-run timeout for astute task

Bug #1539017 reported by Stanislaw Bogatkin
10
This bug affects 2 people
Affects Status Importance Assigned to Milestone
Fuel for OpenStack
Confirmed
Medium
Fuel Toolbox
Mitaka
Won't Fix
Medium
Fuel Toolbox
Newton
Confirmed
Medium
Fuel Toolbox

Bug Description

Today our tasks looks like

parameters:
    cmd: some_cmd
    retries: 10
    interval: 30
    timeout: 180

It means that if cmd returns (for example) 1 as exit code 5 seconds after it ran, it will be restarted after 30 secs again. After 180 secs of such retries task will be stopped.
When solving https://bugs.launchpad.net/fuel/+bug/1533082 we got a problem that cmd in the task itself can hang, so it won't return any exit code at all. In this case task will marked as failed after 180 secs and deployment will be broken.

There is a workaround for this - wrap program in timeout, but it will be much nicer to have such parameter in task itself to ensure that task will be killed and restarted forcibly if it hangs.

tags: added: area-astute area-python
Changed in fuel:
importance: Undecided → Medium
status: New → Confirmed
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.