How to remove UCCX second node from cluster...

[fdetoma]

Hi,

we need to remove subscriber node from UCCX cluster.

We want remain with only primary single node.

How can we do this, in a "correct way"?

I mean,

how can we remove this server in any primary's DB table and/or cluster's settings...?

We must do some task also into UCM...?

If needed, we can reboot primary node.

Thanks!

[Deepak Rawat]

Go to CCX Administration >> System >> Server and delete the entry of secondary server. Post this, remove the VM completely from ESXi host and then restart the UCCX Primary node and that's it. There is nothing else you need to do, when the server is deleted from CCX Administration >> System >> Server, system automatically runs a script in the background that deletes all its associated entries from all the DB tables there is nothing manual you need to do. Till version 8.5(1) SU3 TAC used to run a script by going into root as the deletion process was not very smooth but 8.5(1) SU4 onwards, the above mentioned process takes care of everything automatically.

Regards

Deepak

[fdetoma]

Hi,

I've done this, and now we've only one node.

It works...

thanks!

[Deepak Rawat]

Glad that it worked. Please remember to mark the thread as Answered if there are no further questions on the topic.

Regards

Deepak

[fedor.solovev]

Hi. Please, mark the answer.

[andreast]

Hi,

I am facing a similar but completely different problem:

Installed 11.6 UCCX-cluster. Successfully upgraded publisher to 12.5.

Subscriber already has 12.5 on inactive partition but switch-version failed.

As the system is quite new I thought reinstallation of subscriber might be the easyest way to recover cluster.
Unfortunately trying your instruction to delete subscriber shows: "Error occured while trying to remove node."

Is it finally time for TAC or do you have any hints??

Thx&br

[Roger Kallberg]

Do I understand you correctly that you want to do a new installation of the secondary node and have the database replicate to it or restore it from DRS?  If so you should not remove the second node from the webUI. Doing that will break the possibility to complete the new installation of the second node.

[andreast]

Many thx for your hint! I normally clean up cluster by removing subscriber and recreate it before new setup.
Nevertheless I will follow Kostia's recommendation and contact TAC as I am now interrested in their magic ;)

[oliverpowell]

Question in relation to this comment. I am trying to rebuild the SUB node of a 12.5 cluster. TAC initailly told me to delete the server from the server list in the GUI. I thought that perhaps I would not delete and just rebuild, but changed my mind as the MAC address would change, and I presume it wouldn't restore properly if the MAC is different. So I have tried to delete the server from the gui, but get "Error occured while trying to remove node.".

So currently stumped as what to do!

[Konstantin Vaksin]

Contact Cisco TAC. We can make some magic for you 😊

Kostia

[Anthony Holloway]

TAC can be so awesome to work with.  You guys have saved me several times, and I just love drinking from this coffee mug as a daily reminder that you're there, ready to help in a moments notice.  Thank you!

[Maxim Denisov]

Hello,

Could you tell where this script located or table names where information about nodes located? I can modify DB with dbaccess but can't find  where node configuration located - there are too many tables to check them all. I checked several tables but there is no data about second node.

> select tabname from systables where tabname like "%node%";



tabname  typenodeusage

tabname  typeprocessnoderole

tabname  treecontrolgroupnodes

tabname  processnode

tabname  certificateprocessnodemap

tabname  enterprisenode

tabname  processnodeaudit

tabname  processnodesec

tabname  processnodesecack

tabname  safclientcmnodemap

tabname  processnoderoleservicemap

tabname  processnodeservice

tabname  pickupgroupprocessnodemapdynamic

13 row(s) retrieved.

I have deleted second node from system -> server but still see it in replication servers and see a warning about different ccx and jtapi versions on second node.

Regards,
Maxim

[Roger Kallberg]

Maybe you should create your own post to ask your question as it’s unrelated to the OP topic.