Active IQ Unified Manager Discussions

OnCommand Performance Manager 1.0 question: multiple instance


According OPM 1.0 IAG,  a single OPM instance could support up to 6 nodes cluster.  IHAC has 8-node cluster,  How to deploy multiple instance OPM? in IAG, it has a note:


  1. Attention: Adding the same cluster to more than one instance of Performance Manager impacts the performance of the cluster. When attempting to add a cluster that Performance Manager instance is already monitoring, a warning message is displayed in the GUI.


but I did not find only configure OPM by node rather than by cluster.   How difference OPM instance manage difference nodes?


BTW,  Only 6-nodes are too small for a monitor system.  Any plan to increase the limitation?



Hi Chao,

OPM 1.0 officially supports up to 6-node clusters due to performance limitation of OPM itself. However, we do not put a hard limit in the code to monitor a cluster that has more than 6 nodes. The scale limit is related to the combination of # of clusters, # of nodes, and # of volumes essentially. If your 8-node cluster does not have a large volume count, OPM may be able to handle it. If you add your cluster to OPM and you do not get banners like "collection taking too long" or "analysis taking too long", it will work.

I'm sorry that we do not officially support 8-node clusters in OPM 1.0. But hope the info above can help a little bit.




Hi ruijuan

thank for your kindly reply.

If it's really necessary to deploy multiple OPM instance, is there any guide or document about that?  I did not find useful information in IAG.




Sorry that i forgot to mention that there is no way you can break the nodes to be managed by multiple OPM servers.


Do you mean OPM 1.0 may be not ready for a busy 8+ node cDOT if with many many volumes?   My customer plan to expand to 16 nodes with hundreds volumes. another question if you could help,  if I use  OCUM 6.1+OPM1.0 and OCUM 5.2 to monitor same cDOT,  does it impact cDOT performance?  I read multiple OCUM 6.1 monitoring same cDOT may has impaction.



I just researched and found that we released a newer version of sizing guide in July which increased the supported scale by additional performance improvement and testing in OPM 1.0 GA.

We do support to a single cluster up to 8 node now with volume limit of 500 volumes per node.

This new sizing guide was published in July after the OPM 1.0 GA in May. So there is a discrepancy.


We have qualified up to 8 node cluster during OPM 1.0 with approximately 2000 volumes in total on that cluster. But we do not officially support 8 node since there can be other factors that impact OPM performance. I'm not sure if 16 nodes with hundreds of volumes will work. But it's definitely exceeds the sizing guide limit we provided.

if you use  OCUM 6.1+OPM1.0 and OCUM 5.2 to monitor same cDOT, there is performance impact on cDOT, but should be limited based on my understanding.

NetApp on Discord Image

We're on Discord, are you?

Live Chat, Watch Parties, and More!

Explore Banner

Meet Explore, NetApp’s digital sales platform

Engage digitally throughout the sales process, from product discovery to configuration, and handle all your post-purchase needs.