Travis Nielsen

This appears to affect a number of ceph commands, the operator log also shows `ceph status` and `ceph versions` failing... ``` 2025-05-07 00:26:30.716035 E | ceph-cluster-controller: failed to get ceph...

Note that the `ceph` version inside the rook operator base image is Squid, so perhaps it's a compatibility issue across versions. Rook relies on the `ceph` client compatibility across at...

Fixed already as mentioned previously

1. What is the expected change needed in the orchestrator? Should Rook pass the new `--with-tpm` whenever also passing `--dmcrypt`, if Squid or newer? 2. How will existing encrypted OSDs...

> @travisn should we go ahead with PR and I'll take care of your comment in a new PR? sure, it's not a blocker

@parth-gr Please take a look at the canary test failure, not sure why this change is causing this error for external clusters: ``` The provided rgw Endpoint, '10.101.138.59:80', is invalid....

@parth-gr Thanks for all the troubleshooting, i'll try updating the other images as well to continue troubleshooting.

@parth-gr Could you take another look at this failure? The operator base image, cluster version, and toolbox are all v19, so there should be no differences in the squid version....

I've created a new PR for setting the default ceph version to v19. This PR we can keep solely for the purpose of updating the base image to squid, which...

With the additional commit for logging the specific error code in the external script, we see that the request returns a 403: ``` The provided rgw endpoint, '10.105.220.70:80', is invalid...