CIFS share access slowness with FPolicy EAGAIN errors
Applies to
- ONTAP 9
- CIFS
- FPolicy
Issue
- CIFS client gets severe slowness to access CIFS shares
- The slowness disappears after migrating the data interface to its partner node
- EMS log shows
fpolicy.eagain.on.write
error along withcifsoperationtimedout
errors:
[node1: fpolicy: fpolicy.eagain.on.write:NOTICE]: Write returned EAGAIN while sending notification to the FPolicy server "10.10.10.10" for vserver ID 10.
[node1: kernel: nblade_cifsoperationtimedout_1:ERROR]: params: commandName:SMB2_COM_READ, isQosEnabled:QoS_disabled, localIpAddress:10.10.10.100, lastSpinNpError:SPINNP_ERR_FPOLICY_REQD, clientIpAddress:10.10.10.245, cmdRestartCnt:0, suspensionCnt:2, vserverId:13, system_id:0538184752, sys_version:9.7P11, remoteBladeID:bdba0013-8032-11eb-88d1-d039ea280e22 (node1), lastCsmError:CSM_OK
- Performance Archives might display a high SMB2 Read/Write latency while no latency on the volume layer