Incorrect state of Galera after deployment

[Egor Kotko]

{"build_id": "2014-04-06_01-00-26", "mirantis": "yes", "build_number": "86", "nailgun_sha": "5efa9d21162cd30394f2f608641c324a80ea43dd", "ostf_sha": "17f2fe6e56452f8e2f01a385be4c4b87bf3698a8", "fuelmain_sha": "5c756f65e313144509585768856bfd0869c58d85", "astute_sha": "183fe05cd59a5ce6d154fa263a5a8bf5f27db0ec", "release": "5.0", "fuellib_sha": "9cc637c4f95676928063185ad91b3bed4250c1fe"}

Steps to reproduce:
1. Deploy env:
Centos, 3 Controllers, 1Compute, 1 Cinder, nova FlatDHCP,

Expected result:
Deployed Cluster

 Actual result:
Galera cluster state: http://paste.openstack.org/show/75183/
One of three controllers deployed with errors:

 (/Stage[main]/Galera/Service[mysql]) /etc/puppet/modules/corosync/lib/puppet/provider/service/pacemaker.rb:269:in `stop'
 (/Stage[main]/Galera/Service[mysql]) /etc/puppet/modules/corosync/lib/puppet/provider/service/pacemaker.rb:283:in `status'
 (/Stage[main]/Galera/Service[mysql]) /etc/puppet/modules/corosync/lib/puppet/provider/service/pacemaker.rb:176:in `get_last_successful_operations'
 (/Stage[main]/Galera/Service[mysql]) /etc/puppet/modules/corosync/lib/puppet/provider/service/pacemaker.rb:149:in `get_nodes'
 (/Stage[main]/Galera/Service[mysql]) /etc/puppet/modules/corosync/lib/puppet/provider/service/pacemaker.rb:22:in `get_cib'
 (/Stage[main]/Galera/Service[mysql]) /etc/puppet/modules/corosync/lib/puppet/provider/service/pacemaker.rb:22:in `new'
 (/Stage[main]/Galera/Service[mysql]) /usr/lib/ruby/1.8/rexml/document.rb:43:in `initialize'
 (/Stage[main]/Galera/Service[mysql]) /usr/lib/ruby/1.8/rexml/document.rb:240:in `build'
 (/Stage[main]/Galera/Service[mysql]) /usr/lib/ruby/1.8/timeout.rb:64:in `parse'
 (/Stage[main]/Galera/Service[mysql]) Failed to call refresh: #<REXML::ParseException: #<Puppet::Error: execution expired>

[Vladimir Kuklin (vkuklin)]

According to the logs you have big performance problems. E.g. iptables command (Firewall[204 heat-api]) takes more than 10 seconds. It is the obvious sign that smth is wrong with your environment. Closing as 'Incomplete' until there is a reproducer.

Top lines:
01 - 300.03 sec (5.00 min), line: 6968, Exec[wait-for-synced-state] ERROR!
02 - 300.01 sec (5.00 min), line: 6883, Exec[wait-initial-sync] ERROR!
03 - 300.01 sec (5.00 min), line: 7767, Exec[glance-manage db_sync] ERROR!
04 - 53.01 sec (0.88 min), line: 1363, Package[kmod-openvswitch]
05 - 24.29 sec (0.40 min), line: 2403, Package[libguestfs-tools-c]
06 - 21.19 sec (0.35 min), line: 3030, Package[rabbitmq-server]
07 - 18.75 sec (0.31 min), line: 2696, Package[MySQL-server]
08 - 15.43 sec (0.26 min), line: 3065, Service[rabbitmq-server]
09 - 11.73 sec (0.20 min), line: 2372, Exec[horizon_compress_styles]
10 - 10.96 sec (0.18 min), line: 822, Firewall[204 heat-api]

[Egor Kotko (ykotko)]

There is no performance problems on that host machine.
This bug is duplicated with https://bugs.launchpad.net/fuel/+bug/1303839

[Mike Scherbakov (mihgen)]

What do you mean by "is duplicated with"? If this bug is duplicate of that bug, please mark this one as duplicate - there is a link on this page "Mark as duplicate".