Accepted Solution

Metrocluster: plex naming, aggregate mirroring, FC-VI metrics


We've switched from 3240HA to a Stretch Metrocluster system and I have a few questions.

Namely, I've noticed two things during it's initial configuration:

  1. While creating mirrored plexes of 2 existing aggregates, one was namex plex4 and for another aggregate plex11. I've read somewhere that Data ONTAP names the plexes but I was just wondering where did he come up with those numbers? I see no correlation with neither aggregate nor shelf names.
  2. During initial synchronization (re-sync, level 0) of data on of the aggregates, snapshot reserve was automatically raised to 32%. I was not aware that would happen but i had plenty of room on the aggregate so it was OK. Again, where did Data ONTAP come up with this figure? And more importantly, what would've happened if I didn't have that much space to spare?
  3. Is there a way to monitor data synchronization progress?
  4. Are there some FC-VI metrics available? I'd like to see how it's coping.



Re: Metrocluster: plex naming, aggregate mirroring, FC-VI metrics

1) its just naming it, hidden magic involved, dont bother

2) it will try to grab as much reserve as possible, will drop it after sync. If less space available, it will reserve less space.

3) "sysconfig -r" will show you a resync percentage

4) ic stats error -v

Re: Metrocluster: plex naming, aggregate mirroring, FC-VI metrics

  1. Got it
  2. I thought as much, but worth checking
  3. Weird. I used "aggr status -r" but only saw plex status (re-syncing). Will pay more attention next time.
  4. Ah, there it is... priv set advanced

Thanks, Thomas!