This will help considerably finding out which VMs are affected by some PGs misbehaving.
Description
Description
| Status | Subtype | Assigned | Task | ||
|---|---|---|---|---|---|
| Resolved | MoritzMuehlenhoff | T253824 planned upstream deprecation of the ssh-rsa signing algorithm (RSA with SHA-1) | |||
| Resolved | ayounsi | T254013 all network devices must run OpenSSH >= 7.2p1 but != 7.4p1 | |||
| Resolved | ayounsi | T317175 Junos: resolve DNS through mgmt_junos | |||
| Resolved | ayounsi | T327862 Use mgmt_junos on all network devices | |||
| Restricted Task | |||||
| Resolved | ayounsi | T316539 Upgrade network devices to Junos 20+ | |||
| Resolved | cmooney | T316544 Upgrade cloudsw1-c8-eqiad and cloudsw1-d5-eqiad to Junos 20+ | |||
| Resolved | dcaro | T297083 [ceph] Getting rack level HA | |||
| Open | None | T331636 [cookbooks.ceph] create a script to get the list of rbd images affected by stuck/inactive PGs |
Event Timeline
Comment Actions
I did a first attempt at this, with the current sizes for the chunks that ceph moves around, rbd images are quite spread, and it takes in the order of minutes to get a list of all the placement groups the chunks are in, and for a couple I tried you ended up having most of the fleet (if not all of them), so might not be that useful.
I will not be able to work on it in the foreseeable future, so I'm unassigning it so anyone else can pick it up.