David Galloway

I solved this by deleting the `scheduler-state.json` file. It lived on a Ceph cluster that my docker registry suddenly lost access to and must have corrupted the file.

Can you allow the possibility for using an entire device as swap? On the irvingi, /dev/sdb isn't used at all so I'd like to create a partition on them for...

BTW, looks great so far :+1:

It should but we shouldn't hardcode it in the ceph-cm-ansible.git version. It should be parameterized. So blank definitions here: https://github.com/ceph/ceph-cm-ansible/blob/master/roles/teuthology/defaults/main.yml And our (Sepia's) subnets should be defined here: https://github.com/ceph/ceph-sepia-secrets/blob/master/ansible/inventory/group_vars/teuthology.yml

@b-ranto I'm going through and cleaning up old PRs. I don't understand what this one's for. Is it still needed?

> @djgalloway as this is causing some fuss in other teams, for the soak testing period, can we just make it run for dashboard-labeled PRs? Thanks! Or why don't we...

@jdurgin @neha-ojha are y'all okay with merging this PR? This will disable the `ceph-pull-requests` and `ceph-pr-api` jobs in Jenkins. It will change the GitHub Check names to `ceph PR -...

> @djgalloway something I just realized: and I don't wanna sound like a killjoy, but it'll 'slightly' increment the time taken to finish (10-15 min), as here `make check` unit...

Alright, well, I'm on PTO tomorrow so Monday I guess.

> @djgalloway just started to experiment with pipelines ([ceph/ceph#37265](https://github.com/ceph/ceph/pull/37265)) and it seems it might not require too much rework... Okay. Are we going that route instead, then?