Shotgun executes command one by one instead of doing it simultaneously
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Fuel for OpenStack |
Triaged
|
Low
|
Fuel Sustaining | ||
Mitaka |
Won't Fix
|
Low
|
Fuel Python (Deprecated) | ||
Newton |
Triaged
|
Low
|
Fuel Sustaining |
Bug Description
Steps to reproduce:
1)Modify settings.yaml to have this config for shotgun:
slave:
hosts: []
objects:
- type: command
command: sleep 5 && echo $(date)
to_file: test.txt
2) create snapshot
3) Check the dump.log
cat /var/log/
[node-3.domain.tld] run: sleep 5 && echo $(date)
[node-3.domain.tld] out: Fri Apr 17 04:09:29 UTC 2015
--
[node-1.domain.tld] run: sleep 5 && echo $(date)
[node-1.domain.tld] out: Fri Apr 17 04:09:34 UTC 2015
--
[node-5.domain.tld] run: sleep 5 && echo $(date)
[node-5.domain.tld] out: Fri Apr 17 04:09:39 UTC 2015
--
[node-2.domain.tld] run: sleep 5 && echo $(date)
[node-2.domain.tld] out: Fri Apr 17 04:09:44 UTC 2015
--
[node-4.domain.tld] run: sleep 5 && echo $(date)
[node-4.domain.tld] out: Fri Apr 17 04:09:49 UTC 2015
Command are executed one-by-one, for all nodes.
Shotgun should start executing simultaneously on all nodes.
However, we should carefully test it on big(200+) number of nodes.
Changed in fuel: | |
assignee: | nobody → Fuel Python Team (fuel-python) |
milestone: | none → 6.1 |
Changed in fuel: | |
status: | New → Confirmed |
milestone: | 6.1 → 7.0 |
importance: | Undecided → Medium |
Changed in fuel: | |
assignee: | Fuel Python Team (fuel-python) → Nikolay Markov (nmarkov) |
tags: | added: qa-agree-7.0 |
Changed in fuel: | |
milestone: | 7.0 → 8.0 |
no longer affects: | fuel/8.0.x |
tags: | added: area-python |
I would suggest using parallel-ssh project for this purpose. https:/ /github. com/pkittenis/ parallel- ssh Moving back to team for now.