SnapMirror transfer speed is slow to a FabricPool destination

Applies to

• ONTAP 9.5 and newer with ONTAP Select or Cloud Volumes ONTAP
• ONTAP 9.8 and newer with FAS models

Issue

• When using FabricPool on a destination, such as Cloud Volumes ONTAP, transfer speeds may start out faster but then get slower over time, or may never hit desired throughput.
• Following error messages may be logged in the SnapMirror Audit log when snapmirror fails eventually:

Thu Apr  8 08:43:16 UTC 2021 ResumedUpdate[Apr  7 20:00:00]:081d2038-72d8-11eb-aa21-1f6ad6153aa9 Operation-Uuid=7d7be0a0-ef06-4915-b72c-3090bbab9c0a Group=none Operation-Cookie=0 action=End source=SVM1:vol1 destination=svm_cvo_dr:vol1 status=Failure message=Transfer failed.(Volume access error(No space available in capacity tier staging area))

Failed to start transfer. (Duplicate transfer specified (Replication engine error)

• Following errors can also be seen in Event log just before the Failure:

Tue Nov 08 14:29:27 +1100 [netapp-01: wafl_exempt02: wafl.repl.send.blocks.err:debug]: params: {'SnapMirrorID': '<SnapMirrorID>', 'cpCount': '2079', 'errString': 'Undefined error: 487', 'writeLen': '232144', 'fhFileid': '505', 'errorCode': '487', 'volFsid': '<volFsid>', 'writeOffset': '1792924839936', 'file': 'prod/common/wafl/ng/write.c', 'file_space': 'local metafile', 'volName': '<volName>', 'line': '2376', 'wafl_restarts': '1'}