Endless restart of swift-object-reconstructor and swift-container-reconciler
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Fuel for OpenStack |
Fix Committed
|
High
|
Oleksiy Molchanov | ||
Mitaka |
Fix Released
|
High
|
Oleksiy Molchanov |
Bug Description
Detailed bug description:
The swift-object-
Steps to reproduce:
1. deploy a new environment from the Fuel WebUI
2. after all nodes are up and services have settled, shut down the environment and the fuel master and then start it again (https:/
3. check dmesg for the following repeating messages on the controller nodes:
swift-
swift-
swift-
swift-
Expected results:
container-
or do not start at all.
Actual result:
The swift-object-
When started manually, produce the following errors:
swift@node-1$ /usr/bin/python /usr/bin/
Unable to find container-
swift@node-1$ /usr/bin/python /usr/bin/
Unable to find object-
Workaround:
echo '[container-
echo '[object-
Impact:
1. container-
2. increased CPU consumption due to OS trying to restart services
Description of the environment:
MOS 9.0 build 372
or
MOS 9.0-mos iso 475
OpenStack Release: Mitaka on Ubuntu 14.04
Compute: QEMU
Network: Neutron with tunneling segmentation
Storage Backends: Cinder LVM over iSCSI for volumes
No additional projects/plugins
Total Nodes: 5
Controller + Cinder: 3
Compute: 2
Changed in fuel: | |
assignee: | nobody → Oleksiy Molchanov (omolchanov) |
importance: | Undecided → High |
status: | New → Confirmed |
tags: | added: area-library |
no longer affects: | fuel/newton |
tags: | added: team-bugfix |
description: | updated |
Changed in fuel: | |
status: | Invalid → Confirmed |
description: | updated |
Changed in fuel: | |
status: | Confirmed → In Progress |
Changed in puppet-swift: | |
assignee: | nobody → Oleksiy Molchanov (omolchanov) |
status: | New → In Progress |
no longer affects: | puppet-swift |
tags: | added: on-verification |
I was not able to reproduce your case. These services are enabled on boot, but actually are stopped after boot. Seems that the situation you described happens only if you run 'service $service_name start'