Portal:Cloud VPS/Admin/Runbooks/NodeDownForLong

Overview

The procedures in this runbook require admin permissions to complete.

Error / Incident

A node has been offline (considered down) by a long period of time.

Debugging

This is quite a generic alert, so there's many options, but some things to try:

Try to ssh to the host
Try to ssh to the management port (see SRE/Dc-operations/Hardware_Troubleshooting_Runbook)
Ask on irc if someone is taking any action and forgot about it
Check SAL (or sal.toolforge.org) for host being rebooted/taken down
Check Netbox to make sure the host is not being decommed or network changes were not applied
Check puppet for recent changes that might have affected the node

Common issues

Cloudvirt down

If the host being unresponsive is a cloudvirt host, it will trigger a page. You should add it to the "maintenance" aggregate to prevent new VMs being scheduled on that host, and drain the host evacuating all VMs to other hosts.

There is a cookbook wmcs.openstack.cloudvirt.drain, but it does not work if the host is unreachable, tracked in phab:T375223.

Example:

root@cloudcontrol1005:~# openstack aggregate add host maintenance cloudvirt1063

# take note of the status of all VMs (ACTIVE, SHUTOFF, PAUSED), as you'll have to manually restore it after the VM is moved to another host
root@cloudcontrol1005:~# openstack server list --host cloudvirt1063 --all-projects

# the value for os-password can be found in /etc/novaadmin.yaml
root@cloudcontrol1005:~# nova --os-username novaadmin --os-project-name admin --os-auth-url "https://openstack.eqiad1.wikimediacloud.org:25357/v3" --os-password XXXXXX --os-user-domain-id default host-evacuate cloudvirt1063
nova CLI is deprecated and will be removed in a future release

+--------------------------------------+-------------------+---------------+
| Server UUID                          | Evacuate Accepted | Error Message |
+--------------------------------------+-------------------+---------------+
| 3585384e-e85f-41da-a014-448c59d76586 | True              |               |
| abaa359b-3468-4b56-be7c-24dc88252c9e | True              |               |
| 5899163e-bbfb-474e-8a9a-2dae5006de22 | True              |               |
| f9015696-9010-4a1c-a7cf-9bc8e8098df4 | True              |               |
| 15700b39-5773-4fec-9349-47f16bc457bd | True              |               |
| b615abe0-9cc2-4cee-a224-c1f0b90e7957 | True              |               |
| 20f22f9d-5734-49b9-8a19-9908e588c92f | True              |               |
| 1b0a01e4-e8cb-4a68-ad6c-306b9869db05 | True              |               |
| 3192a8d0-0fda-4d24-a1b7-2e2dc20fbe0e | True              |               |
+--------------------------------------+-------------------+---------------+

# restart VMs that were in ACTIVE status with:
root@cloudcontrol1005:~# openstack server start <id>

Remember to manually resolve the alert in VictorOps, otherwise it will fire again after 24 hours.

Related information

Support contacts

Communication and support

Support and administration of the WMCS resources is provided by the Wikimedia Foundation Cloud Services team and Wikimedia movement volunteers. Please reach out with questions and join the conversation:

Discuss and receive general support

Chat in real time in the IRC channel #wikimedia-cloud ^connect or the bridged Telegram group
Discuss via email after you have subscribed to the cloud@ mailing list

Stay aware of critical changes and plans

Subscribe to the cloud-announce@ mailing list (all messages are also mirrored to the cloud@ list)
Read the News wiki page

Track work tasks and report bugs

Use a subproject of the #Cloud-Services Phabricator project to track confirmed bug reports and feature requests about the Cloud Services infrastructure itself

Read stories and WMCS blog posts

Read the Cloud Services Blog (for the broader Wikimedia movement, see the Wikimedia Technical Blog)

Old incidents

Add your incident here: