-
Bug
-
Resolution: Done
-
Major
-
ACM 2.12.0
-
False
-
None
-
False
-
Provide a fallback xattr implementation for cases where this is unsupported by the storage being used for the assisted PV
-
-
-
None
Description of problem:
Using the ACM / Multicluster engine to create new cluster, the creation process gets stuck on the "Preparing" step of the hosts. The hosts are successfully discovered, and shows Ready, but when the installation starts it goes through the "Preparing for install" and gets to "Preparing step successful" for all hosts and gets stuck there. This worked fine on 4.16, but trying two different hub clusters on 4.17.2 and 4.17.4 and both fails the same way (with different set of hosts).
Version-Release number of selected component (if applicable):
4.17.4
How reproducible:
Start an installation of a new OpenShift cluster, using a file system for assisted service that doesn't support extended attributes (like RHEL NFS before 8.4 or NetApp ONTAP NFS before 9.12), go through the discovery of the hosts and continue on with the installation
Steps to Reproduce:
0. Ensure that assisted service is using a file system that doesn't support extended attributes. 1. Start the installation of a new cluster through the ACM UI 2. Complete the discovery of the hosts (either with discovery ISO or PXE boot) 3. Fill in the necessary information and click next to start the installation of the cluster
Actual results:
Installation never completes
Expected results:
Installation works and a new cluster is operational
Additional info:
We have attempted multiple ways of discovering the hosts - i.e.: discovery ISO, PXE boot, etc. Have also attempted to create a Host Inventory first before starting the install - all methods are failing the same way. Please let us know what other information may be needed - this issues is very easily reproduced in our environment.
- clones
-
ACM-16240 [ACM-2.13] Provisioning failure due to lack of extended attributes
-
- Closed
-
- links to