Cisco Unified MeetingPlace, Release 7.0 -- How to Perform Application Server Failover

From DocWiki

Main page: Cisco Unified MeetingPlace, Release 7.0

Up one level: Configuration

Contents 1 Performing Application Server Failover 1.1 Before You Begin 1.2 Procedure 1.3 Verifying 1.4 Troubleshooting Tips 1.5 Related Topics 2 Performing Application Server Failover to Switch to the Previous Active Server 2.1 Scenario 1: Voluntary Failover 2.1.1 Before You Begin 2.1.2 Procedure 2.1.3 Verifying 2.2 Scenario 2: Required Failover 2.2.1 Before You Begin 2.2.2 Procedure 2.2.3 Verifying 2.3 Troubleshooting Tips 2.4 Related Topics

Performing Application Server Failover

Before you perform this task:

• Node 1 is the current active server.

• Node 2 is the current standby server.

After you complete the task:

• Node 1 will be the standby server.

• Node 2 will be the active server.

Before You Begin

Set up Application Server failover by completing one of the following sections:

• Setting Up Failover for Two Newly Installed Application Servers

• Setting Up Failover for One Existing Application Server and One Newly Installed Application Server

Caution! Performing this task temporarily brings down your Cisco Unified MeetingPlace system. Perform this task only if the current active server fails, or during a scheduled maintenance period.

Procedure

1. If the Node 1 is up, complete the following steps:
   1. Log in to the CLI of Node 1.

      If you are logging in remotely, use the Node 1 eth0:0 IP address or hostname.

   2. Enter the following command to change the server to standby mode:

      failoverUtil setServer standby

2. Log in to the CLI of Node 2.

   If you are logging in remotely, use the Node 2 eth0:0 IP address or hostname.

3. Enter the following command to change the server to active mode:

   failoverUtil setServer active

Verifying

Using the hostname or IP address of the virtual eth0:0 interface, log in to the Administration Center of each node, and verify that the correct failover deployment mode (active or standby) appears at the top of the page.

Troubleshooting Tips

If the Administration Center displays the wrong failover mode on either or both servers, something may have interrupted the process initiated by the failoverUtil setServer command. To resolve this issue, see the failoverUtil setServer command description.

Related Topics

• Using the Command-Line Interface (CLI) in Cisco Unified MeetingPlace

• Performing Application Server Failover to Switch to the Previous Active Server

Performing Application Server Failover to Switch to the Previous Active Server

This section describes two scenarios for performing application server failover.

Scenario 1:Voluntary Failover

• Node 2 is the active server.
• Node 1 is the standby server and working fine, but you need to failover to Node 1 for some reason.

Scenario 2: Required Failover

• Node 2 is the active server.
• Node 1 is the standby server, but was removed from replication (either due to hardware failure or some other reason).

After you complete the task for either scenario:

• Node 1 will be the active server.
• Node 2 will be the standby server.

Scenario 1: Voluntary Failover

Before You Begin

Caution! Performing this task temporarily brings down your Cisco Unified MeetingPlace system. Perform this task only if the current active server fails, or during a maintenance period.

• This procedure assumes the following scenario:
  • Node 2 is the active server.
  • Node 1 is the standby server and working fine, but you need to failover to Node 1 for some reason.

• Complete Prerequisites for Application Server Failover.

• Back up and archive the data on the currently active Application Server (Node 1). See Backing Up, Archiving, and Restoring Data on the Cisco Unified MeetingPlace Application Server.

Procedure

1. Sign in to the CLI of Node 2.

   If you are signing in remotely, use the Node 2 eth0:0 IP address or hostname.

2. Enter the following command to change the server to standby mode:

   failoverUtil setServer standby

3. Sign in to the CLI of Node 1.

   If you are signing in remotely, use the Node 1 eth0:0 IP address or hostname.

4. Enter the following command to change the server to active mode:

   failoverUtil setServer active

5. If Node 2 is down, enter the following command to switchOFF replication from Node-1:

   mp_replication switchOFF –r node2-eth0:0-hostname

Verifying

Using the hostname or IP address of the virtual eth0:0 interface, log in to the Administration Center of each node, and verify that the correct failover deployment mode (active or standby) appears at the top of the page.

Scenario 2: Required Failover

Before You Begin

Caution! Performing this task temporarily brings down your Cisco Unified MeetingPlace system. Perform this task only if the current active server fails, or during a maintenance period.

• This procedure assumes the following scenario:
  • Node 2 is the active server.
  • Node 1 is the standby server, but was removed from replication (either due to hardware failure or some other reason).

• Complete Prerequisites for Application Server Failover.

• Back up and archive the data on the currently active Application Server (Node 1). See Backing Up, Archiving, and Restoring Data on the Cisco Unified MeetingPlace Application Server.

Procedure

1. If Node 1 was brought to single-server mode, perform these steps:
   1. Sign in to Node 2 using eth0:0 to bring Node 2 into Standby mode:

      failoverUtil setServer standby

   2. Sign in to the Node 1 CLI using the console.
   3. Enter the following command to set up failover:

      failoverUtil setDeployment failover

   4. Follow the CLI prompts to configure the virtual network interface (eth0:0) with an IP address, subnet mask, default gateway, and hostname.

      Node 1 automatically restarts and enters standby mode.

   5. Enter the following command to initialize database replication:

      mp_replication init -s 1 -n 1 -l node1-eth0:0-hostname -r node2-eth0:0-hostname -h

   6. Enter the following command to synchronize existing data from Node 2 and start database replication:

      mp_replication switchON -r node2-eth0:0-hostname -S -F node2-eth0:0-hostname

2. If Node 1 was never brought to single-server mode, but database replication configuration was removed from Node 1, do the following:
   1. Sign into the CLI of Node 1. If you are signing in remotely, use the Node 1 eth0:0 IP address or hostname.
   2. Enter the following command to initialize database replication:

      mp_replication init -s 1 -n 1 -l node1-eth0:0-hostname -r node2-eth0:0-hostname -h

   3. Enter this command to synchronize existing data from Node 2 and start database replication:

      mp_replication switchON -r node2-eth0:0-hostname -S -F node2-eth0:0-hostname

3. Sign in to the CLI of Node 2.
4. Enter the following command to change Node 2 to standby mode:

   failoverUtil setServer standby

5. Sign in to the CLI of Node 1.
6. Enter the following command to change Node 1 to active mode:

   failoverUtil setServer active

Verifying

Using the hostname or IP address of the virtual eth0:0 interface, log in to the Administration Center of each node, and verify that the correct failover deployment mode (active or standby) appears at the top of the page.

Troubleshooting Tips

If the Administration Center displays the wrong failover mode on either or both servers, something may have interrupted the process initiated by the failoverUtil setServer command. To resolve this issue, see the failoverUtil setServer command description.

Related Topics

• Using the Command-Line Interface (CLI) in Cisco Unified MeetingPlace

• Performing Application Server Failover