cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1638
Views
0
Helpful
3
Replies

Dbreplication 7.1.5 fails, clusterreset fails, everything fails!

carlnewton
Level 3
Level 3

Guys,

I noticed on a 7.1.5 cluster i have at the moment that DBreplication is not working right.  Whatever i did, the replication setup fails.

I tried utils dbreplication stop on all subs, then on the pub, then did utils dbreplication clusterreset from the pub but it doesnt work.

I did a fresh reboot of ALL servers, utils dbreplication runtimestate shows all servers with rpc up from the pub. Doing the same from the subs does not show all servers but replication has never worked to them so i think they dont know about each other.

After doing utils dbreplication stop on all subs, then the pub when thats complete, i run utils dbreplication clusterreset and it fails, with the logs showing this over and over again:

15:33:52 Deleting server g_iHOSTNAME_ccm7_1_5_30000_1 from the sub.
15:33:56 Deleting server g_iHOSTNAME_ccm7_1_5_30000_1 from the pub.
15:33:56 Done deleting  g_iHOSTNAME_ccm7_1_5_30000_1 trying to define again.
15:33:56 Defining Subscriber iHOSTNAME_ccm7_1_5_30000_1
15:34:03 Publisher cdr list check
15:34:04 Starting Loop of cdr list check
15:34:04 Listing the subscriber found g_iHOSTNAMEB_ccm7_1_5_30000_1
15:34:34 Define of Sub g_iHOSTNAME_ccm7_1_5_30000_1 did not work, deleting and trying again.
15:34:34 Deleting server g_iHOSTNAME_ccm7_1_5_30000_1 from the sub.
15:34:36 Deleting server g_iHOSTNAME_ccm7_1_5_30000_1 from the pub.
15:34:37 Done deleting  g_iHOSTNAME_ccm7_1_5_30000_1 trying to define again.
15:34:37 Defining Subscriber iHOSTNAME_ccm7_1_5_30000_1
15:34:43 Publisher cdr list check
15:34:44 Starting Loop of cdr list check
15:34:44 Listing the subscriber found g_iHOSTNAMEB_ccm7_1_5_30000_1
15:35:14 Define of Sub g_iHOSTNAME_ccm7_1_5_30000_1 did not work, deleting and trying again.
15:35:14 Deleting server g_iHOSTNAME_ccm7_1_5_30000_1 from the sub.

I have replaced the actual hostname of the servers with HOSTNAME. Note there is also a HOSTNAMEB in there (two diff servers)

It seems to fail to define the first sub in the list and gets no further.  Interestingly the server is called "HOSTNAME" and not "g_iHOSTNAME" but i dont know if this is an issue or not.

All subs are defined with IP Addresses on the CCM.  I dont really know where to go from here with this, normally clusterreset fixes everything!

Does anyone have anything else i can try on this?

3 Replies 3

gogasca
Level 10
Level 10

Can you run a Database Status Report via Unified Reporting tool to get more details.

Can you perform nslookup and reverse lookup between all nodes in the cluster?

Any DNS change?

Any firewall in the middle?

Thanks

I got this resolved in the end with some help.

The rhosts file on one of the nodes was wrong and the CDR sys tables were wrong.

a restart of the A cisco DB service on the affected node resolved the Rhosts file.

we then ran a utils dbreplication stop on all nodes, followed by a utils dbreplication dropadmindb on all nodes.

once this was complete a utils dbreplication reset all resolved the problem.

thanks anyway!

Great to know that its working fine now.

You should always start by checking the hosts,rhosts and sqlhosts files on the CUCM servers to check whether they match. If they don't a restart of the A Cisco DB and a utils dbreplication stop on the affected node and on the pub and utils dbreplication start usually helps.

Check the following doc about troubleshooting DBReplication issue

https://supportforums.cisco.com/docs/DOC-13672

regards

Nitesh