-
Bug
-
Resolution: Cannot Reproduce
-
Undefined
-
None
-
4.17
-
Quality / Stability / Reliability
-
False
-
-
None
-
Important
-
None
-
None
-
None
-
None
-
None
-
Customer Escalated
-
None
-
None
-
None
-
None
-
None
-
None
-
None
Description of problem:
When running a massive migration of virtual machines from VMWare to OCP Virtualization (MTV), the multiple storage operations done by the process appears to hit some limitation on the Dell's storage array that slow down the time on where virt-launcher pods in OCP takes to start up the virtual machine. Migrating a VM takes a very long time as the required Volumes spend a lot of time to be properly created/mounted.The array is struggling because of the many attach/detach map/unmap operations.
Version-Release number of selected component (if applicable):
OCP 4.17.15 Baremetal PowerEdge R7625 CSI Driver for Dell EMC PowerMax/v2.12.0+79 MTV operator v2.7.11
How reproducible:
Intermittently
Steps to Reproduce:
Initiate several virtual machine migration through the MTV operator Some VMs will migrate quickly, other will get stuck waiting for the volumes, provisiones by the DELL CSI drivers, to be presented.
Actual results:
Virtual machine migration from VMWare to OCP-V takes long time to complete due to a delay in the volume provisioning.
Expected results:
- Will need to check from the OCP storage side if we have any issues that might contribute to the behavior on mounting volumes and/or adding to the consumed time to do so. - Even though the workload is related to OCP-V/MTV, the main issue is not directly related to OCP-V, we believe that this behaviour might also happen with other workflow that involves many PVs (Live Migration, Cluster upgrades, draining nodes, etc)
Additional info: