-
Notifications
You must be signed in to change notification settings - Fork 1.9k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
node-drain using name #863
Comments
btw, if someone is having the same problem, or this issue gets closed because you consider that is not neccesary, workaround here:
|
@adrianlop Thanks for filing this. This may be a documentation issue as it has always been by UUID. The reason it has to be UUID is that the name doesn't necessarily have to be unique. So it would be ambiguous which node needed draining! Can you point out the documentation that shows using a name |
Yeah, I saw it here: Also, here it says node ID or prefix: I understood node0* by prefix, I tried it but no good result neither: I see now that if 2 nodes have these 2 UUIDS:
So it's no problem, if you think this is not a valuable enhancement you can close it! |
There was one small piece in @adrianlop 's original issue that I don't want to be overlooked: a |
@dadgar Thanks! |
I'm going to lock this issue because it has been closed for 120 days ⏳. This helps our maintainers find and focus on the active issues. |
Hi,
Upgrading our Nomad cluster to 0.3, I encountered a problem with the node-drain command.
In the example of the documentation it seems that you can specify a node using its name (
$ nomad node-drain -enable node1
), but when I run the command, it doesn't work. It only accepts UUID:Error toggling drain mode: Unexpected response code: 500 (rpc error: rpc error: node lookup failed: index error: UUID must be 36 characters)
I think it is useful that you can drain a node specifying the name (I'm using Ansible to automate this upgrade, and since the nomad client name is the same as the hostname, it would be really easy to achieve this).
It would be useful too, to launch the command with a
myself
param or similar, so only that node where you launched the command gets drained. This way you don't have to write the UUID or the name, and simplifies the automation of rolling upgrades using a CM like Ansible/puppet etc.What do you think?
Thank you guys!
The text was updated successfully, but these errors were encountered: