While upgrading 1241 SNOs OCP platform from 4.10.32 to 4.11.5 and Operators from v4.10 to v4.11, the TALM operator "precached" on a preparation CGU and consumed 42 minutes of time presumbly precaching nothing as the disk usage of the actual SNOs did not increase.

Deployed SNO OCP 4.10.32
Upgraded to 4.11.5

Ran during this one time, however I assume it is rather easy to reproduce.

1.
2.
3.

The CGU didn't become precaching complete until almost a full 43 minutes later

Stats on clustergroupupgrades CRs in namespace ztp-platform-upgrade-prep
Total CGUs - 1
CGUs with PrecachingCompleted: 1 - 100.0%
CGUs with UpgradeCompleted: 1 - 100.0%
Earliest CGU creationTimestamp: 2022-10-26 19:06:40
Latest PrecachingDone lastTransitionTime: 2022-10-26 19:49:35
Earliest CGU startedAt Timestamp: 2022-10-26 19:50:36
Latest CGU completedAt Timestamp: 2022-10-26 20:05:14
Duration between creation and startedAt: 2636.0s :: 0:43:56
Duration between creation and precachingdone: 2575.0s :: 0:42:55
Duration between startedAt and completedAt: 878.0s :: 0:14:38
Duration between creation and completedAt: 3514.0s :: 0:58:34

If there is nothing to precache, no precache jobs should be created on the spoke clusters listed by the CGU.