Need to add automated case for checking connectivity after ban l3-agent many times

[Georgy Dyuldin]

We need automation of next scenario "Ban l3-agent many times and check l3-agent work":

Steps:

Create network1, network2
Create router1 and router2, connect networks to the external net
Boot vm1 in network1 and associate floating ip
Boot vm2 in network2 and associate floating ip
Add rules for ping
ping vm1 and vm2 from each other with floatings ip
Ban 2 l3-agents: pcs resource ban p-neutron-l3-agent node-_node_id
Create sh-file: touch name_of_the_file.sh
Write script which ban and clear one agent about 40 times
Run the script with a parameter: bash name_of_the_file.sh node-<number>.domain.tld
Wait some time
Check that l3-agent on the ban|clear node is alive: neutron agent-list
Boot one more VM (VM3) in network1
Go to VM1 console in Horizon
ping 8.8.8.8
ping VM3 by internal IP and by floating IP
Go to VM3 console in Horizon
ping 8.8.8.8
ping VM1 by both IP, ping VM2 by floating IP

which should check North-South, East-West connectivity and situations with and without floatings:
http://paste.openstack.org/show/482703/

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/254695

[Roman Prykhodchenko (romcheg)]

Georgy, please attach world accessible links to the bug description.

[OpenStack Infra (hudson-openstack)]

Change abandoned by Georgy Dyuldin (<email address hidden>) on branch: master
Review: https://review.openstack.org/254695