Open Stack

We included in all of our puppet manifest a tools directory that has a "restartworld.sh" [1].  We specify classes of servers [App, network, compute] which run classes of services and a specific service that we wish to restart or if you want to restart everything you do "world".  Then we have an ansible playbook [2] that will  run restartworld.sh using the class of server automatically.  So if you want to restart everything you just tell it to run against all hosts in an environment with a service of world and it will do everything for you.

[1] - https://github.com/godaddy/openstack-puppet/blob/master/tools/restartworld.sh
[2] - https://github.com/godaddy/openstack-ansible/blob/master/playbooks/restartworld.yaml

The following links are currently a bit old - we revamped them to work with systemd and our cells setup.  I will work on getting those updated onto our public repo.
____________________________________________

Kris Lindgren
Senior Linux Systems Engineer
GoDaddy, LLC.

From: David Medberry <openstack at medberry.net<mailto:openstack at medberry.net>>
Date: Tuesday, March 17, 2015 at 9:39 AM
To: "openstack-operators at lists.openstack.org<mailto:openstack-operators at lists.openstack.org>" <openstack-operators at lists.openstack.org<mailto:openstack-operators at lists.openstack.org>>
Subject: [Openstack-operators] Restart Services when Rabbit Croaks

Hi all,

In PHL Mid-Cycle Operators meetings, several folks in the rabbitmq session indicated they basically restart all openstack services when they have a rabbit issue. Can you please share your process/scripts with this list.

At the present time, with an Ubuntu Trusty base OS, I'm doing:


sudo restart libvirt-bin ; sudo restart nova-compute ; sudo restart neutron-plugin-openvswitch-agent ; #ALL COMPUTE NODES
sudo restart neutron-server ; sudo restart neutron-l3-agent; sudo restart neutron-dhcp-agent ; sudo restart neutron-metadata-agent ; sudo restart neutron-plugin-openvswitch-agent ;sudo restart nova-scheduler ; sudo restart nova-api ; sudo restart nova-cert; sudo restart nova-conductor; sudo restart nova-consoleauth; sudo restart nova-novncproxy ; sudo restart nova-objectstore ; #ALL CONTROL NODES

and sometimes

sudo restart ceilometer-collect ; sudo restart ceilometer-agent-notification ; sudo restart ceilometer-agent-central ; sudo restart ceilometer-api #ALL CONTROL NODES

for SVC in cinder-api        cinder-backup     cinder-scheduler  cinder-volume; sudo restart $SVC; #All CONTROL NODES


(or the equivalent via ansible)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-operators/attachments/20150317/b1cd7cbe/attachment.html>