Travis Nielsen

@maxpain Did you apply this setting on the pool? It is not enabled by default. ```yaml enableCrushUpdates: true ``` For example, see [here](https://github.com/rook/rook/blob/master/deploy/examples/pool.yaml#L36-L37)

The [Ceph EC docs](https://docs.ceph.com/en/latest/rados/operations/erasure-code/) indicate that the erasure code profile cannot be updated, which includes the crush rule. Therefore, Rook does not update the ec profile or failure domain for...

Ok interesting that you were able to update the EC pool with a new profile. You are seeing the PGs move around as expected to respect the new failure domain,...

@maxpain Good to hear it appears to be working, though I don't expect to automate it in rook since ceph doesn't support it.

According to @rzarzynski it's not supporting in ceph, so I'm guessing this falls in the category of "do at your own risk, it might work, but it's not tested".

@zsksy123 The issue is that the OSDs are not up and running. Only one is up, so there was some issue starting the OSDs. ``` osd: 3 osds: 1 up...

If you create a new cephobjectstore does it succeed? Or you could delete that cephobjectstore and then re-create it. Perhaps since it was in the failing state while OSDs down,...

Ok great to hear it is working now, there must be an issue to recover automatically from that failure. I should have asked sooner, but can you perhaps share the...

The solution to this issue is really to have a per-cluster operator, as specified in https://github.com/rook/rook/issues/13848. It keeps coming up for various issues like this, so we are certainly looking...

Yes, definitely planning to look into it, and trying to find time to prototype it soon...