Fuel for OpenStack

Controller and compute went offline after provisioning

Bug #1322175 reported by Andrey Sledzinskiy on 2014-05-22

This bug affects 2 people

Affects		Status	Importance	Assigned to	Milestone
	Fuel for OpenStack	Invalid	High	Andrey Sledzinskiy	Fuel for OpenStack 5.1

Bug Description

reproduced on {"build_id": "2014-05-22_01-10-31", "mirantis": "yes", "build_number": "216", "ostf_sha": "5c479f04c35127576d35526650ec83b104f9a33d", "nailgun_sha": "31fcb161ff8d6bfb861b041467440752c0e9c537", "production": "docker", "api": "1.0", "fuelmain_sha": "dda9cff27999e74b80e8e6a9e665e7e9677ab994", "astute_sha": "9a0d86918724c1153b5f70bdae008dea8572fd3e", "release": "5.0", "fuellib_sha": "872b6a7a968b619a493ad46c504910020ea2edae"}

Steps:
1. Create cluster - Centos, simple, KVM, Flat Nova Network, Ceph for volumes and for images
2. Add 1 controller, 1 compute, 2 ceph nodes
3. Click deploy cluster

Actual result - after provisioning Controller and compute stop responding and went offline after timeout

Logs are attached

Revision history for this message

Andrey Sledzinskiy (asledzinskiy) wrote on 2014-05-22:

fuel-snapshot-2014-05-22_13-12-44.tgz Edit (9.9 MiB, application/x-tar)

Revision history for this message

Dmitry Borodaenko (angdraug) wrote on 2014-05-22:

Looks like an IO error on the CI host (out of disk space maybe?), observe node-1.test.domain.local/kernel.log:

2014-05-22T09:40:54.539904+00:00 crit: EXT4-fs error (device dm-0) in ext4_reserve_inode_write: IO failure
2014-05-22T09:40:54.539904+00:00 crit: EXT4-fs error (device dm-0): file system corruption: inode #393554 logical block 0 mapped to 1644040 (size 7)
...
2014-05-22T09:40:54.540692+00:00 crit: EXT4-fs error (device dm-0): file system corruption: inode #675346 logical block 1606 mapped to 191046 (size 6)