Hi,

They should be always running, but there are some prior issues I think that might stop them on an upgrade or restart. One workaround can be found here (depending on what code you are on):

The default performance monitors may pause after a GeoSynchrony upgrade or when the VPLEXManagementServer console restarts.

To restart the perpetual monitoring, run the following command :

VPlexcli:/>ll /monitoring/directors/*/monitors/

This should be run on both mgmt stations if you have a metro and see the issues.

After a little while you should see the log files with "perpetual" in their name or you can run a collect-diagnostics and see if that grabbed them as well.

The following CLI command generally speaking is the best first option to try:

          VPlexcli:/> debug engineering-mode on

     VPlexcli:/> emc-internal configuration restart-perpetual-monitors

          VPlexcli:/> debug engineering-mode off

5.4 SP1 specific, this KB mentions the "ll /monitoring/directors/*/monitors/" command, but there is some more to it and is only 5.4 specific:  https://support.emc.com/kb/334191

As for how long you need to get a good view of performance - well, that all depends but generally speaking I'd say at least 24 to 48 hours is required to get a good feel for performance.  Perpetual monitor files contain about 60 days or so worth of history.

Instantaneous performance over the last 5 minutes is visible via the Unisphere GUI.

Be sure to also check out the VPM (VPLEX Performance Monitor) for virtual-volumes which can store I believe up to 30 days of per virtual-volume information.

Gary