[fuel-qa][venv]Provision of node with multipath device fails on boot

[Alexander Zatserklyany]

Steps to reproduce:
-------------------
1. Setup Fuel with multipath device and 3 slave nodes.
2. Create environment with 1 controller, 1 compute and 1 cinder.
3. Provision nodes.

Expected results:
-----------------
All nodes a provisioned.

Actual results:
-----------------
Controller node can't boot.

The boot process fails with the following messages:
 Incrementally starting RAID arrays...
 mdadm: CREATE user root not found
 mdadm: CREATE group disk not found
 Incrementally starting RAID arrays...
and slowly repeats the above at this point.

This issue appears only with qemu-kvm virtual environments.

[Sergey Slipushenko (sslypushenko)]

Put Importance in to High because this bug appears only on the virtual environments

[Aleksey Zvyagintsev (azvyagintsev)]

Issue not reproducible.
 Waiting for result from additional set of fuel-qa tests.

[Aleksey Zvyagintsev (azvyagintsev)]

After deeper investigation, we found that currently devops\fuel-qa spawn 'unsuitable' VMs for multipath emulation,
so, test venv should be improver to support :

1)All slaves on 'raw' disks
2)Disable snapshoting for slaves (since raw not support snapshots
3) Add to VM xml new section :
node_xml.controller(type='scsi', model='virtio-scsi')
4) Add posibility to switch-off disk caching (under testing)

[Igor Shishkin (teran)]

@Alexey, fuel-devops team is not in charge for python-devops and/or tests.
Assigning to QA team accordingly.