Error on second controller with Ubuntu cluster

[Nastya Urlapova]

{"build_id": "2014-04-05_04-17-26", "mirantis": "yes", "build_number": "268", "nailgun_sha": "7a05e365240ab27c492b20585ef8ac8557102cc0", "ostf_sha": "de0222fed646525d248dc6892eeceab139d5c469", "fuelmain_sha": "5714f2a3cf56ebdc124e9555dfff0fe274e84bdd", "astute_sha": "55df06b2e84fa5d71a1cc0e78dbccab5db29d968", "release": "4.1", "fuellib_sha": "52e7f57695f33bafa5d84d524d77f1bc3a2289b2"}[

Cluster - Ubuntu in HA 3xController and 2XCompute

err: (/Stage[corosync_setup]/Osnailyfacter::Cluster_ha::Virtual_ips/Cluster::Virtual_ips[management_old]/Cluster::Virtual_ip[management_old]/Cs_commit[vip__management_old]/cib) change from absent to vip__management_old failed: Execution of '/usr/sbin/crm_shadow --force --commit vip__management_old' returned 186: Could not commit shadow instance 'vip__management_old' to the CIB: Communication error on send

[Mike Scherbakov (mihgen)]

If whole HA in Ubuntu doesn't work, then it should be Critical, not High, I assume... this is our core feature.

[Vladimir Kuklin (vkuklin)]

this bug is sporadic and, moreover, puppet retry will finish deploy of the 2nd node.

[Nastya Urlapova (aurlapova)]

This other issue from #1303839

[Vladimir Kuklin (vkuklin)]

never seen a reproducer for this bug. We need to confirm it as it looked like simple connectivity issue.

[Drazic Lin (drazic-lin)]

when I tried these steps, I got the same error logs.

mode: HA
compute:qemu
network:neutron w/ gre
cinder: default
glance: ceph

deploy with 2 controllers + 1controller &ceph + 1ceph

[Dmitry Ilyin (idv1985)]

I have deployed Ubuntu HA on our current master without such error, neither it's reported by our CI. Logs show not only cib commit error but also failure to assemble galera. Perhaps connectivity issue?

Anyway we dont see such problems on our current code base...