controller goes offline if cluster is actively used
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Mirantis OpenStack |
Invalid
|
Undecided
|
Dina Belova |
Bug Description
[root@fuel ~]# fuel --fuel-version
api: '1.0'
astute_sha: f7cda2171b0b677
auth_required: true
build_id: 2015-01-20_08-49-23
build_number: '38'
feature_groups:
- mirantis
fuellib_sha: 9aa913096fb93ea
fuelmain_sha: 1ee1766a51bdb5b
nailgun_sha: 5f91157daa6798f
ostf_sha: 3d2f44dcfa32d6c
production: docker
release: 6.0.1
release_versions:
2014.2-6.0:
VERSION:
api: '1.0'
astute_sha: f7cda2171b0b677
build_id: 2015-01-20_08-49-23
build_number: '38'
feature_
- mirantis
fuellib_sha: 9aa913096fb93ea
fuelmain_sha: 1ee1766a51bdb5b
nailgun_sha: 5f91157daa6798f
ostf_sha: 3d2f44dcfa32d6c
production: docker
release: 6.0.1
Ubuntu, HA, neutron-gre, Ceilometer, debug, Ceph for volumes,
controllers: 3
computes: 97
Controller hangs when we perform some load (rally tests).
Console of one of controller nodes unreachable via ssh and even via ipmi. We seen node-47 is hanged last time.
If we reboot the hanged controller, we can see the following behaviour:
https:/
diagnostic snapshot is here https:/
Changed in mos: | |
assignee: | nobody → MOS Keystone (mos-keystone) |
no longer affects: | fuel |
Changed in mos: | |
assignee: | MOS Keystone (mos-keystone) → arogusskiy (arogusskiy) |
Changed in mos: | |
milestone: | none → 6.1 |
Changed in mos: | |
assignee: | arogusskiy (arogusskiy) → Dina Belova (dbelova) |
it's baremetal installation