System tests 7.0 doesn't worked because of problem with libvirt

[Vadim Rovachev]

https://patching-ci.infra.mirantis.net/view/All/job/7.0.system_test.ubuntu.bonding_ha/11/console

+ python fuelweb_test/run_tests.py -q --nologcapture --with-xunit --group=bonding_ha
Create environment and set up master node ... libvirt: Storage Driver error : Storage pool not found: no storage pool with matching name 'default'
libvirt: Storage Driver error : Storage pool not found: no storage pool with matching name 'default'
libvirt: Storage Driver error : Storage pool not found: no storage pool with matching name 'default'
libvirt: Storage Driver error : Storage pool not found: no storage pool with matching name 'default'
libvirt: Storage Driver error : Storage pool not found: no storage pool with matching name 'default'
libvirt: Storage Driver error : Storage pool not found: no storage pool with matching name 'default'
libvirt: Storage Driver error : Storage pool not found: no storage pool with matching name 'default'
libvirt: Storage Driver error : Storage pool not found: no storage pool with matching name 'default'
libvirt: Storage Driver error : Storage pool not found: no storage pool with matching name 'default'
libvirt: Storage Driver error : Storage pool not found: no storage pool with matching name 'default'

[Igor Shishkin (teran)]

Jan 18 10:18:58 srv031-scc puppet-agent[10975]: Prefetching virsh resources for libvirt_pool
Jan 18 10:18:58 srv031-scc puppet-agent[10975]: Executing '/usr/bin/virsh -q pool-list --all'
Jan 18 10:18:58 srv031-scc puppet-agent[10975]: (/Stage[main]/Fuel_project::Jenkins::Slave/Libvirt_pool[default]) Dependency File[/etc/sysconfig/libvirtd] has failures: true
Jan 18 10:18:58 srv031-scc puppet-agent[10975]: (/Stage[main]/Fuel_project::Jenkins::Slave/Libvirt_pool[default]) Skipping because of failed dependencies

[Igor Shishkin (teran)]

The issue is fixed on those servers, but we still need to check why it wasn't passed from the first time(second puppet pass created the pool).
Decreasing priority to medium.

[Vadim Rovachev (vrovachev)]

Thank you so much, Igor.

[Andrey Nikitin (heos)]

CI-team, please provide a server to test deployement of jenkins slave by manifests.

[Aleksandra Fedorova (bookwar)]

mc0n1-msk provided

[Andrey Nikitin (heos)]

As I can see, the target pool was created:

virsh # pool-list
 Name State Autostart
-------------------------------------------
 default active yes

Moreover, I can't find related errors in the log files. So, I think, that problem was happeden once.

[Andrey Nikitin (heos)]

I've tried to deploy clean server with our puppet manifests and I see, that the problem doesn't repeat.

Related job on jenkins sandbox:
https://jenkins-sandbox.infra.mirantis.net/job/7.0.system_test.ubuntu.bonding_ha_one_controller/5/console