Storage Replication Adapter: Long running operations due to inefficient QOS Policy Group parsing
- Views:
- 85
- Visibility:
- Public
- Votes:
- 0
- Category:
- ontap-storage-replication-adapter
- Specialty:
- virt
- Last Updated:
Applies to
- ONTAP Tools for VMware vSphere (OTV) 9.8
- ONTAP Tools for VMware vSphere (OTV) 9.10
Issue
When running a VMware Site Recovery Manager (SRM) operation that triggers a device discovery, the following is observed:
- Operation takes longer than expected
- A bulk of the discoverDevices operation is spent on parsing flexvol QOS Policy Groups
For example, in the vvolvp.log, the following events show an example of where QOS Policy Groups are being evaluated. Based on the number of flexvols with QOS Policy Groups assigned, this parsing may take an extended period of time and encompass hundreds if not thousands of events (lines chopped for readability):
2022-04-08T14:24:19,761Z (DEBUG)[utilsPool-3681-thread-1] [] Sending-1564816: <qos-policy-group-get-iter><max-records>5000</max-records></qos-policy-group-get-iter>
2022-04-08T14:24:19,849Z (DEBUG)[utilsPool-3681-thread-1] [] Received-1564816: <results status='passed'><attributes-list><qos-policy-group-info><is-shared>true</is-shared><max-throughput>
2022-04-08T14:24:19,849Z (DEBUG)[utilsPool-3681-thread-1] [] Retrieve of zapi qos-policy-group-get-iter took 88 milliseconds for iter #1
2022-04-08T14:24:19,849Z (DEBUG)[utilsPool-3681-thread-1] [] retrieve: end
2022-04-08T14:24:19,849Z (DEBUG)[utilsPool-3681-thread-1] [] parse: begin
2022-04-08T14:24:19,850Z (DEBUG)[utilsPool-3681-thread-1] [] parse: end, groups = {Extreme_prod_mir=QosPolicyGroupImpl{name=Extreme_prod_mir, max throughput=6144iop