-
Bug
-
Resolution: Unresolved
-
Normal
-
CNV v4.18.0
-
Quality / Stability / Reliability
-
0.42
-
False
-
-
False
-
CNV v4.99.0.rhel9-2292
-
-
CNV I/U Operators Sprint 268, CNV I/U Operators Sprint 269, CNV I/U Operators Sprint 271, CNV I/U Operators Sprint 272, CNV I/U Operators Sprint 273, CNV I/U Operators Sprint 274, CNV I/U Operators Sprint 275, CNV I/U Operators Sprint 276, CNV I/U Operators Sprint 277, CNV I/U Operators Sprint 278
-
Moderate
-
None
Description of problem:
The metric for kubevirt_vmi_migration_failed include VMIMs in 'undefined' status
Version-Release number of selected component (if applicable):
4.18.0.rhel9-813
How reproducible:
100%
Steps to Reproduce:
1. Migrate 1000s of VMs 2. Monitor migration metrics and compare with data captured with oc get vmim
Actual results:
Failure metric does not match number of failed VMIMs
Expected results:
Failure metric matches number of failed VMIMs when queried with oc get vmim Separate metric for undefined state, along with metrics for Scheduled and TargetReady.
Additional info:
Metrics: sum(kubevirt_vmi_migration_failed) sum(kubevirt_vmi_migrations_in_pending_phase) Attached screenshot shows an inverse relationship between "failed" and "pending", when in-fact VMIMs show 0 failed, and a relative number of "undefined" state VMIMs.
- is cloned by
-
-
- Closed
-
- is related to
-
CNV-61079 [CLOSED LOOP for] The metric for kubevirt_vmi_migration_failed include VMIMs in 'undefined' status
-
- Closed
-
- is triggered by
-
CNV-54728 Performance upgrade testing for 4.18.0
-
- Closed
-
- relates to
-
-
- Closed
-
-
CNV-56998 RN: Known issue - VirtualMachineInstanceMigrations failing when combined pending and undefined migrations approaches 400
-
- Closed
-
- mentioned on
