VPlex: monitor collect command fails with error "Firmware returned unrecognized statistic 'fe-init ops |'. This is a bug."

The creation of custom Vplex performance monitors related to frontend statistical collection generates an error. It is specific to the following host initiator performance stats that were introduced in Geosynchrony 6.x

Host initiator performance stats new to Geosynchrony 6.x

host-init.ops                                         initiator-ports    counter         counts/s

host-init.ops-act                                     initiator-ports    reading         counts

host-init.read                                        initiator-ports    counter         KB/s

host-init.read-avg-lat                                initiator-ports    period-average  us

host-init.unmap-avg-lat                               initiator-ports    period-average  us

host-init.unmap-ops                                   initiator-ports    counter         counts/s

host-init.write                                       initiator-ports    counter         KB/s

host-init.write-avg-lat                               initiator-ports    period-average  us

Sample commands to illustrate the issue: 

To create-

monitor create -p 5min -n director-1-1-A_host-init-stats -d director-1-1-A -s host-init.* -t /clusters/cluster-1/exports/initiator-ports/*

To add sink file - 

monitor add-file-sink -m director-1-1-A_host-init-stats -f /var/log/VPlex/cli/reports/director-1-1-A_host-init-stats.csv

To manually trigger data collection -

monitor collect -m director-1-1-A_host-init-stats

The error - 

monitor collect: Evaluation of <<monitor collect -m director-1-1-A_host-init-stats>> failed.
cause: Failed to collect data.
cause: Firmware returned unrecognized statistic 'fe-init ops |'. This is a bug.

Workaround: 

When creating the monitors, ensure that you exclude unregistered initiator ports. You can do this by using a wildcard (*) value that excludes unregistered initiators.  See the example listed below.

Alternatively, if you register all of the unregistered initiators this will also work around the issue. (navigate to "/clusters/cluster-x/exports/initiators"  and use the "register" command as per the Vplex CLI guide)


Example: 

VPlexcli:/> ll /clusters/cluster-1/exports/initiator-ports/

/clusters/cluster-1/exports/initiator-ports:
Name                           port-wwn            node-wwn            type          Target Port Names
-----------------------------  ------------------  ------------------  ------------  --------------------------
HOST1_HBA0                     0x5001438028xxxxxxx  0x50014380xxxxxx  default
HOST1_HBA1                     0x5001438021xxxxxxx  0x50014380xxxxxx  default
HOST2_HBA0                     0x500143801xxxxxxxx  0x50014380xxxxxx  default
UNREGISTERED-0x10000000222xx  0x1000000022222xxx  0x20000000c33xxx  default       P0000000046E006E4-A0-Fxxx,
                                                                                     P0000000046F00xxx-B0-FC01
UNREGISTERED-0x50012481044444xxx  0x50012481044444xx  0x5001248112121xxx  RecoverPoint

Based on the above output, you would need to create monitors using the following syntax: 

monitor create -p 5min -n director-1-1-A_host-init-stats -d director-1-1-A -s host-init.* -t /clusters/cluster-1/exports/initiator-ports/HOST*

This will include all the initiators that are registered  (starting with "HOST") and will exclude the unregistered initiators.