Selective High-Availability with Two Cisco Prime DCNM Appliances

The below figure shows a setup in which High Availability has been selectively enabled for auto-configuration feature. This means that failover is available only for LDAP. Node-1 is configured as active with all the applications necessary for Fabric deployments. This includes DCNM, LDAP, XMPP, AMQP, TFTP, DHCP. Node-2 is configured as standby with only LDAP (necessary for auto-configuration). With the initial configuration, DCNM on Node-1 uses LDAP on Node-1 for updates. The standby LDAP (on Node-2) gets synced with the active LDAP (on Node-1) via sync replication.

When active LDAP (on Node-1) fails, DCNM automatically uses the standby LDAP (on Node-2) to make any further updates, thus providing application failover.

Once selective HA is configured on Node-1, backups for all applications are triggered at regular intervals (by default, every six hours) and transferred to Node-2 and stored at

/var/lib/backups/selective-ha/active_backups

Once selective HA is configured on Node-2, the LDAP is backed-up at regular intervals (by default, every six hours) and transferred to Node-1 to be stored at

/var/lib/backups/selective-ha/standby_backups

On both nodes, only the last 5 backups are maintained.

Note You can change the backup interval. See Changing the Backup Interval.

Selective High-Availability with Redundancy

To have redundancy and to reduce the downtime in case of a server failure, it is advisable to deploy a redundant pair of active and standby nodes during the initial setup. The sequence would be:

• Setup Selective HA for the first pair of servers. For more information, see Selective High-Availability with Two Cisco Prime DCNM Appliances.
• Setup the second pair with an identical network configuration as the secondary server. Ensure that the IP address and port groups assigned to the interfaces of the primary server must be the same as the secondary server.
• In case of server failure of any of the hot servers, power on the corresponding spare server and restore the backup from the hot server.

Time synchronization using NTP server

For most of the HA functionality to work, you must synchronize the time on both appliances by using an NTP server. The installation would typically be in the management access network (eth0) interfaces.

Configuring Active Cisco Prime DCNM Appliance

Step 1 Log in to the SSH terminal of the Cisco Prime DCNM Appliance of the Active appliance and enter the following command:

appmgr setup selective-ha active –feature auto-config

Step 2 Make sure that each prerequisite is in place and press y if not all of the pre-requisites are in place, press n to exit.

A prompt for the management IP of the standby peer appears.

After you enter the management IP of the standby appliance, the active appliance generates a pair of authentication keys and transfers it to the peer’s authorized keys.

Step 3 Enter the IP address of the NTP server and press y to continue.

After the active appliance is configured for Selective High-Availability, the following message appears:

Step 4 Verify the HA role of the node by using the appmgr show ha-role command.

Configuring Standby Cisco Prime DCNM Appliance

Step 1 Log into the SSH terminal of Standby Cisco Prime DCNM appliance and enter the appmgr setup selective-ha standby-feature auto-config command.

Step 2 Press y to continue.

The details of the active Cisco Prime DCNM appliance will be available in the standby, using which the standby appliance generates a pair of authentication keys and transfers it to the Cisco Prime DCNM appliance’s authorized keys.

Carefully check if it is the correct Cisco Prime DCNM appliance and press y to continue. Once the active appliance is configured for Selective High-Availability, and the following message is displayed.

**********************************************************************************

**********************************************************************************

Step 3 Check the HA role of the node by entering the appmgr show ha-role command.

Starting Applications in Active Cisco Prime DCNM Appliance

Step 1 Login into the SSH terminal of the Active Cisco Prime DCNM appliance.

Step 2 Start all applications by using the appmgr start all command.

Wait for all the applications to start.

Starting Applications in Standby Cisco Prime DCNM Appliance

Step 1 Login into the SSH terminal of the Standby Cisco Prime DCNM appliance and start the LDAP applications using the appmgr start ldap command.

Changing the Backup Interval

By default, on a node that has selective HA enabled, backup is performed every six hours. You can change the backup interval by following the below steps.

Step 1 Log into the SSH terminal of the node on which you want to change the backup interval and open the crontab in a vi editor.

Step 2 Edit the entry for the script selective-ha-backup.sh.

For example, the below configuration triggers a backup every hour.

Step 3 Load the crontab configuration file using the command

crontab /etc/crontab

Load the crontab configuration file into crontab

Step 4 Verify the cron entry by running

crontab -l

Recovering DCNM Active Node

When Cisco Prime DCNM hot active node fails, you need to bring up the redundant spare active node, which has an identical configuration in terms of IP address and HA configuration as the original setup. (In a 4 box setup, you already have this ready, and need to likely power on the appliance to have this usable).

Step 1 Log into the SSH terminal of Standby node and navigate to the backup directory which contains the backups from the active node.

Standby-peer# cd /var/lib/backups/selective-ha/active_backups

Step 2 Identify the most recent active backup based on timestamp and move it to the spare DCNM active node.

Standby-peer#sftp root@x.x.x.x

x.x.x.x> put backup.mm_dd_yyyy__HH_MM_ss.tar.gz

Step 3 Log into the SSH terminal of DCNM active node and verify that all the applications are up and running by using the following command:

appmgr status all

Step 4 After all the applications are up, stop the applications by using the following command:

appmgr stop all

Step 5 Restore the backup using the following command:

Active-peer# appmgr restore all backup. mm_dd_yyyy__HH_MM_ss.tar.gz

Step 6 Once the restore is completed, start all applications on active node by using the appmgr start all command.

Step 7 Log into the SSH terminal of Standby node and restart the LDAP

Standby-peer# appmgr restart ldap

Recovering DCNM Standby Node

When the DCNM hot standby node fails, bring up the spare standby node, which has an identical configuration in terms of IP address and HA configuration as the original setup. (In a 4 box setup, you already have this ready, and need to likely power on the appliance to have this usable).

Step 1 Log into the SSH terminal of active node and navigate to the backup directory location which has the backups taken on the standby node.

Active-peer# cd /var/lib/backups/selective-ha/standby_backups

Step 2 Identify the most recent standby backup based on timestamp and move it to the spare DCNM standby node.

Standby-peer#sftp root@x.x.x.x

x.x.x.x> put backup.mm_dd_yyyy__HH_MM_ss.tar.gz

Step 3 Log into the SSH terminal of spare DCNM standby node and verify if the LDAP application is up and running by using the following command:

appmgr status ldap

Step 4 After the LDAP application is up, stop the application by using the following command:

a ppmgr stop ldap .

Step 5 Restore the LDAP application backup by using the following command.

Standby-peer# appmgr restore ldap backup. mm_dd_yyyy__HH_MM_ss.tar.gz

Step 6 After the restore operation is completed, start the LDAP application on standby node by using the following command:

appmgr start ldap

Step 7 Log into the SSH terminal of Standby node.

Step 8 Restart the LDAP application by using the following command:

Active-peer# appmgr restart ldap