CUCM Subscriber with different OS password from publisher

Unanswered Question
Sep 4th, 2010
User Badges:

Hi,


i have a CUCM cluster (publisher and subscriber) version 7.1.3ASU1 and we added two days ago another subscriber but during the installation it only ask for administrator password. Now i have my new subscriber with same application user name and password but different OS password. Does this have any issue for the replication?? please let me know how to change the password for the OS to be the same as the old servers.



regards,

Ibrahim Chehouri

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Loading.
David Hailey Sat, 09/04/2010 - 08:53
User Badges:
  • Purple, 4500 points or more

So, if I read you correctly:


What you should have is the Pub and the Sub share the Application Administrator and password (e.g., ccmadministrator) which logs you into the CCM Admin interfaces.  This user/password are replicated.


The Platform Administrator is what most commonly refer to as the OS Administrator.  It's generally named something like "administrator" or "admin".  So, you set the Platform Administrator account to the same name but you set the Platform Administrator password differently for the 2 servers.  This doesn't cause an issue as the Platform user/password are not replicated.  So, you can continue with this set up if you want.  If not, it is safe to change the password as well.  Log in to the CLI on the server you want to change the Platform Admin password for and enter:


set password user admin


You will be prompted to enter the old password and then enter the new password twice.


That's it.


Hailey


Please rate helpful posts!

ichehouri Sat, 09/04/2010 - 11:03
User Badges:

Hi Hailey,


yes iam talking about Platform user name and password. i am having a strange behaviour, on the publisher server i can see only two servers (the old servers) on the Database status report with good status and on the old subscriber i see also two servers (the subscribers only) with good status while on the new subscriber i can see the three servers (all of them) with a good status for old servers and initializing for the new subscriber.


note that any change on the publisher is replicated on the other servers directly without delay which means they are synchronizing.


any suggestions please.



regards,

Ibrahi

David Hailey Sat, 09/04/2010 - 11:48
User Badges:
  • Purple, 4500 points or more

Did you reboot all servers after adding the new node? If not, that's a required step. If you did then you may need to reset DB replication for the cluster.


Sent from my iPhone

ichehouri Sat, 09/04/2010 - 14:00
User Badges:

i rebooted only two servers (publisher and the new subscriber). i did not reboot one server (subscriber) because users was connected. i will try to do it tonight.if it still have the same issue, how can i reset DB replication for the cluster?




regards,

Ibrahim

David Hailey Sat, 09/04/2010 - 17:56
User Badges:
  • Purple, 4500 points or more

You should've rebooted all the servers in the cluster first. So, let's do that before we go any future.  Reboot all servers starting with pub.


Hailey


Rate my post!

ichehouri Mon, 09/06/2010 - 01:42
User Badges:

i rebooted all the servers yesturday and i am still having the same issue. Note that i followed the documents for adding the new node.

the publisher database status (under reporting) is just showing two servers (the old ones) while on the new subscriber database status i can see all the servers with initializing for the new node



any suggestions.


ibrahim

David Hailey Mon, 09/06/2010 - 06:54
User Badges:
  • Purple, 4500 points or more

You will likely need to reset the replication within the cluster at this point.


1) On each subscriber, enter the CLI command: utils dbreplication stop

2) On the publisher, enter the CLI command: utils dbreplication stop (do not do this before you have stopped replication on the subscribers first)

3) On the publisher, enter the CLI command: utils dbreplication reset all


This will reset DB replication on all servers in the cluster may take a while to complete.  So, I'd say give yourself 45 minutes or so and then check to see what the status of replication is and that you can see all servers in the host table on each node.


If your problem persists after that, you may have a problem elsewhere.


Hailey

Actions

This Discussion