Create volume and attach it to instance test failed on Ubuntu.BVT
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Fuel for OpenStack |
Fix Released
|
High
|
Aleksandr Didenko |
Bug Description
VERSION:
feature_groups:
- mirantis
production: "docker"
release: "6.1"
api: "1.0"
build_number: "97"
build_id: "2015-02-
nailgun_sha: "62dd6289785079
python-
astute_sha: "ed5270bf9c6c12
fuellib_sha: "df615cc9206126
ostf_sha: "c9100263140008
fuelmain_sha: ""
Ubuntu.BVT failed with error:
Create volume and attach it to instance (failure): Time limit exceeded while waiting for volume becoming 'in-use' to finish.
Test Result:
http://
Scenario:
1. Create cluster
2. Add 3 nodes with controller and ceph OSD roles
3. Add 1 node with ceph OSD roles
4. Add 2 nodes with compute and ceph OSD roles
5. Deploy the cluster
6. Check ceph status
Duration 90m
Snapshot ceph_ha
I've got similar bug on custom BVT with absolutely the same simptoms.:
http:// jenkins- product. srt.mirantis. net:8080/ view/custom_ iso/job/ custom. ubuntu. bvt_2/393/
I've managed to debug broken env for some time and it looks like monit is the root cause here. So here's what's happenning on compute node during deployment:
1) Puppet configures and starts monit service
Wed Feb 04 11:19:52 +0000 2015 Puppet (debug): Executing '/etc/init.d/monit start'
2) Monit starts 'nova-compute' since it's not running yet:
[UTC Feb 4 11:19:53] info : 'nova-compute' start: /usr/sbin/service
2) Puppet finishes configuration for ceph and nova-compute and evalutes 'nova-compute' service: nova-compute] (info): Starting to evaluate the resource nova-compute] (info): Evaluated in 0.06 seconds
Wed Feb 04 11:20:01 +0000 2015 /Service[
Wed Feb 04 11:20:01 +0000 2015 /Service[
But it's already started by monit.
So it looks like monit starts nova-compute whichs interfere with deployment process where this service should be configured and started by puppet.