Guest

Cisco Unity Connection

Unity Connection Cluster Replication Rebuild

Techzone Article content

Document ID: 116942

Updated: Dec 14, 2013

Contributed by Scott Hills and Brad Magnani, Cisco TAC Engineers.

   Print

Introduction

This document describes steps to verify and attempt to address the issue when replication becomes out of sync or breaks entirely.

Problem

It is important to know that there are two types of replication that occur within Unity Connection (UC):

  • Enterprise Replication (ER) - platform Cisco CallManager (CCM) related
  • Unity Connection Replication

Enterprise Replication

It is important that ER is always Real Time Monitoring Tool (RTMT) state of (2)'s on both the Publisher and the Subscriber. 

  • In order to confirm this, enter the utils dbreplication runtimestate command.
  • Another helpful command that can be used is utils dbreplication status.

Unity Connection Replication

Without the proper ER running, Unity Connection Replication does not work correctly. Commands used to troubleshoot UC replication are:

  • utils cuc cluster overwritedb
  • utils cuc cluster renegotiate

Solution

If ER works (RTMT states = 2) but UC replication is down, complete these steps:

  1. On the Subscriber, enter the utils dbreplication stop command. Wait for it to complete before you start step 2.
  2. On the Publisher, enter the utils dbreplication stop and utils dbreplication reset all commands.

If ER is down (RTMT does not = 2), complete these steps:

  1. On the Subscriber, enter the utils dbreplication stop  command. Wait for it to complete before you start the next step.
  2. On the Publisher, enter the utils dbreplication stop command. Wait for it to complete before you start the next step.
  3. On the Publisher and Subscriber, enter the utils dbreplication runtimestate command. Ensure that both servers are RPC reachable column = YES).
  4. On the Publisher, enter the utils dbreplication dropadmindb command.
  5. On the Subscriber, enter the utils dbreplication dropadmindb command.
  6. On the Publisher, enter the utils dbreplication clusterreset command. Wait for it to complete before you start the next step. If you run Release 9.x, skip to step 8.
  7. Restart the Subscriber. Wait for the Subscriber to come back up and services all start before you start the next step.
  8. On the Publisher, enter the utils dbreplication reset all command.
  9. On Publisher and Subscriber, periodically enter the utils dbreplication runtimestate command in order to monitor the RTMT state they each have for one another. They should progress and both ultimately end up at (2) if replication sets up properly. This can take some time.
  10. If RTMT states do not go from 0's to 2's after you wait a good amount of time, collect this information from BOTH the Publisher and the Subscriber:

    • utils dbreplication runtimestate
    • file get activelog cm/trace/dbl/*.log
    • file get activelog cm/trace/dbl/sdi/dbmon*.txt
    • file get activelog cm/log/informix/dbl_repl*.log
    • file get activelog cm/log/informix/ccm*.log
  11. Open a case with the Technical Assistance Center (TAC) and provide the information collected in step 10.

Related Information

Updated: Dec 14, 2013
Document ID: 116942