Travis Nielsen

You have changed the total number of chunks from 9 --> 6. Previously, the data chunks were 7, so now the data appears to be in a state where it...

> Swapping back to 7 from 6, has not returned the filesystem to a working state. What were the pairs of `dataChunks + codingChunks` (`k + m`) that you've set...

@rzarzynski Any ideas on recovering from this assert when trying to change the EC params?

When the CephBlockPool EC parameters are updated, Rook will update the EC profile with a command such as this, which will force the update: ``` ceph osd erasure-code-profile set ec-pool_ecprofile...

@rzarzynski Thanks for the clarification. Looking again, I don't see a requirement where we were trying to support updating the EC params, it was just a side effect of some...

To see the settings, try: ``` ceph osd erasure-code-profile ls ceph osd erasure-code-profile get ``` If Rook isn't reconciling your EC settings due to the health errors, then you can...

Not sure, sounds like the OSDs are failing before they get a chance to remove the PGs for the EC pool. @rzarzynski Any workaround to get the OSDs back up...

@sethjones To try using `ceph-objectstore-tool`, you'll want to put an OSD in [maintenance mode](https://github.com/rook/kubectl-rook-ceph/blob/master/docs/maintenance.md), then run the needed commands while an OSD is offline.

Sounds good to add this option to the cluster chart. I assume this would be configured in the values similar to the [current ingress](https://github.com/rook/rook/blob/release-1.17/deploy/charts/rook-ceph-cluster/values.yaml#L416-L439). @onedr0p Are you interested in opening...

@ayashjorden The changes will be included in the v1.17.7 release, which will be out in the next 2-3 days. Please wait until then.