Chapter 6. Bug fixes

Multicloud Object Gateway is making API requests to the incorrect URL when OpenShift Data Foundation is deployed on Microsoft Azure Government (MAG)

Previously, the default backing store was faulty, and the overall system health was degraded. This is because the Multicloud Object Gateway (MCG) creates a default backing store on the platform where OpenShift Data Foundation was deployed. The default backing store was created with the wrong values due to an issue in identifying the Microsoft Azure Government (MAG).

With this update, a Persistent Volume (PV) pool is created when identifying the MAG instead of creating the usual Azure backing store, and later on, you can create a backing store on top of the MAG environment and use that new backing store.

(BZ#1944572)

The default backing store 'noobaa-default-backing-store' is stuck in the 'Creating' state

Previously, the Cloud Credentials Operator (CCO) changed credentials after the creation of the default backing store. Due to this, the backing store used to get stuck. The update credentials on the default backing store are fixed with this release. The default Multicloud Object Gateway (MCG) backing store does not get stuck on creating while changing credentials.

(BZ#1992090)

S3 listing goes into an infinite loop, when the prefixes are of names 99 and 990

Previously, the query results for the listing operation were not in the order that the Multicloud Object Gateway (MCG) expected them to be as the collate algorithm was set incorrectly while starting the MCG DB.

With this update, the collate algorithm is set correctly on new deployments while starting the MCG DB. Now, the query results are in the expected order.

(BZ#2068110)

Usage of MD5 in a FIPS-enabled environment is explicitly allowed and S3 multipart operations can be completed

Previously, in a FIPS-enabled environment, the usage of MD5 digest was not allowed by default, unless explicitly excluded for non-cryptographic purposes. Due to this, a segfault occurred during the S3 complete multipart upload operation.

With this fix, the usage of MD5 for non-cryptographic purposes in a FIPS-enabled environment for S3 complete multipart PUT operations is explicitly allowed and the S3 multipart operations can be completed.

(BZ#2086724)

OpenShift Data Foundation pods are restarting due to liveness probe failure

Previously, the liveness probe on pods caused a restart of Ceph pods. This release update increases the default timeout for the liveness probe. The pods now get more time before restarting due to liveness where the nodes have more loads and fewer CPU/memory resources.

(BZ#2091951)

OSD pods are in CrashLoopBackOff state after the upgrade to OpenShift Data Foundation 4.10.3 failed to deactivate the volume group for the local volume

Previously, LVM-based OSDs from older OpenShift Data Foundation releases (4.3, 4.4) were incompatible with version 4.10 and would crash when upgraded to 4.10. With this update, the OSDs from older OpenShift Data Foundation releases are running instead of crashing. Rook now adds the correct flag to handle these legacy OSDs so that they will be functional with 4.10.

(BZ#2097268)