CIDuty/How To/Take actions to RelEng Hardware from TaskCluster UI

Info

• Taskcluster team have added hardware control actions into taskcluster interface for the releng hardware machines
• They have added options to ping and reboot the machines running on linux, windows and osx.

Use case

1. Check to see what client ID you have (check Manage Credentials page, accessible from the Taskcluster Account menu)
2. Activate VPN.
3. Read about roller and authorize the self-signed cert (internal only so they don't need certificate for that) using this. Accept the security issue/warning and proceed to add an exception for it.
4. Go to the hardware you want to take actions.

The page should look like this:

• Alongside the Quarantine button, we have Ping and Reboot buttons. Go ahead and try one of it (ensure that the hardware has all the tasks completed, don’t restart a machine that has running tasks unless that’s the whole point you want to do it).
  

• If everything is setup correctly, by pressing the ping button, it will ping the machine and after a few seconds you will receive an email with the ping result.
  

On IRC channel #roller, a message will be sent by a bot, containing the action that has done and the LDAP account of the user who initiated the action.

• Same goes for the reboot, but takes much more (aprox. 60-80 seconds).
  

Response Example

Known problems

• If you try to restart it without being logged on and/or without VPN, an error will appear when you press one of the options.
• The result/notification of that action is send to the person who initiates it, via email and IRC but, it tries to guess your LDAP email and IRC , for example, if our usernames would be for example “test_name”, he would say that on IRC, your username is test_name and LDAP is test_name@mozilla.com.
• On some cases, that is a problem because the LDAP user and IRC name is not the same, therefore you will only get the email notification.