ActiveIQ 09.8P1 Unified Manager : Cluster polling does not finish

[StockageUGA]

Hi all,

We have reinstalled a Fresh ActiveIQ 9.8 P1 (Vmware) !

Added connections to 5 clusters, First polling took a lonnnngg time ! And one out of 5 clusters is still not finished (after 24 hours)

Where can I look for details/debug/... ?

Thanks,

GS.

[pedro_rocha]

Hello my friend,

You must generate a support bundle and try the ocumserver.log under <support bundle name>\server directory.

You can generate the support bundle from the Maintenance Console (there is also an option for a smaller bundle: "Generate Light Support Bundle").

Regards,

Pedro

[StockageUGA]

Hi,

Managed to get the bundle. ocumserver.log is quite long to read 😄

Searched for "error" or the clustername, only found one line to investigate :

2021-03-09 08:58:31,604 ERROR [oncommand] [cluster-ems-configure-3] [c.n.dfm.impl.ems.EMSConfigureUtil] Communication status of Cluster ********** is not good. Can't configure EMS.

And by the way, It seems that the polling just finished like just right now !

Storage Mgmt -> Cluster Setup  (additionnal columns) :

First line is our problematic cluster ! Compared to the others, seems that It takes too much time.

Regards,

GS

[pedro_rocha]

Yeap, that's really strange.

Do all the clusters run the same ONTAP version? Same model? Any network difference? Network Topology are the same for all?

You could try to capture network packets to understand what could be happening...

I did not find any bugs or related info in the release notes. Maybe you should open a case with NetApp...

Regards,

Pedro

[StockageUGA]

Cannot be that simple 😉

Different ONTAP versions (9.7/9.8), different hardware (FAS/AFF), different buildings, different routers, different firewalls, and so on...

We will keep monitoring it, we reinstalled ActiveIQ from scratch (old one was successive upgrade from OCUM7) because polling was taking too much time on another cluster...

more than 6PB of data... dozens of SVMs, hundreds volumes/qtrees,  100thousands users quotas,  takes time I guess !

Thanks,

GS.

[pedro_rocha · Accepted solution]

Definitely a big deployment...

If the cluster taking too long is a big one, maybe could be those many ONTAP structures... even though seems too long taking  that much time to complete...

Smells networking... 😛

Please post here if you sort it out...

good luck!