We would like a nagios plugin script that would check all the following:
Is https://koji.fedoraproject.org/koji answering and returning reasonable looking content?
Is https://koji.fedoraproject.org/kojihub answering and returning some test api call content?
Is https://kojipkgs.fedoraproject.org/ answering and returning some package content?
If any of these fail, return a error state to nagios with a message on why.
Hello, is anyone working on this issue? I would like to work on it as my first contribution :)
Not that I know of. Please take it and work on it.
Metadata Update from @nb: - Issue assigned to lrossetti
Any news here @lrossetti ?
Metadata Update from @kevin: - Issue priority set to: Waiting on Asignee - Issue tagged with: monitoring
Attaching the patch file.
I was able to test it on my local nagios instance but I didn't use the ansible playbooks so I may need some help with regards to e2e testing.
<img alt="6505.patch" src="/fedora-infrastructure/issue/raw/files/1ed702471d357b82652b6feb92afdab151efb8183940fffbbaec8735f0c00362-6505.patch" />
Please refer to the ansible-6505.patch attachment instead of the "6505.patch" one.
<img alt="ansible-6505.patch" src="/fedora-infrastructure/issue/raw/files/d5fefa47faa2a84546cb33d1c92f46c3c5a024fb6ba08cb8618ca8382f8e6b50-ansible-6505.patch" />
I appologize for how long it's taken to get back to this. :frowning:
This plugin looks pretty great!
In calling the check though we need to probibly use koji.fedoraproject.org instead of koji01.phx2.fedoraproject.org so we can make sure and test the endpoint on the proxies that everyone hits (and could go to koji01 or koji02).
Otherwise it looks great. You want to change that in the patch before I apply, or shall I just do so after I apply it?
No worries, I took a while to work on it as well :)
Attaching the updated patch file (ansible-6505-2.patch)
<img alt="ansible-6505-2.patch" src="/fedora-infrastructure/issue/raw/files/bcf67b148cdf83f5760be855cbcdb34656a35e28a3e18f633d8f31436d3a2064-ansible-6505-2.patch" />
I applied this, but we need more adjustment for the hostname change I'm afraid.
koji.fedoraproject.org isn't a valid host on the internal nagios. :(
I reverted it back until you or I have more time to look.
@kevin what do you think if we add a default var to store a list of koji hostnames?
Well, I don't think it will work for all of them... but I guess yeah, we could for those that it would work fine. (all of the names koji has that is).
Metadata Update from @cverna: - Assignee reset - Issue tagged with: backlog
Metadata Update from @cverna: - Issue assigned to cverna
So I went ahead and just changed the hostname in the service definition (https://infrastructure.fedoraproject.org/cgit/ansible.git/commit/?id=82208b3d42de914ddf23333fc44fd6230608b23a)
That way the plugin still checks for koji.fedoraproject.org hitting the proxy (http requests) but nagios reports any problem with this service on the koji01.fp.o host.
koji.fedoraproject.org
This is now deployed and running. I ll close this ticket but we can reopen it if the above solution does not work.
Metadata Update from @cverna: - Issue close_status updated to: Fixed - Issue status updated to: Closed (was: Open)
Commit 96662a3a fixes this issue
Login to comment on this ticket.