Cisco Prime Infrastructure – Configuring High Availability

An HA deployment consists of two Prime Infrastructure servers: a primary and a secondary. Each of
these servers has an active database and a standby backup copy of the active database. Under normal circumstances, the primary server is active: It is connected to its active database while it manages the
network. The secondary server is passive, connected only to its standby database, but in constant
communication with the primary server.

 

The Health Monitor processes running on both servers monitor the status of its opposite server. Oracle
Recovery Manager (RMAN) running on both servers creates the active and standby databases and
synchronizes the databases when there are changes, with the help of Oracle Data Guard Broker running
on the primary server.

 

When the primary server fails, the secondary takes over, connecting to its active database, which is in
sync with the active primary database. You can trigger this switch, called a “failover”, either manually,
which is recommended, or have it triggered automatically, You then use the secondary server to manage
the network while working to restore access to the primary server. When the primary is available again,
you can initiate a switch (called a “failback”) back to the primary server and resume network
management using the primary.

 

reference : Cisco.com

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

In any Prime Infrastructure HA implementation, for a given instance of a primary server, there must be
one and only one dedicated secondary server.

 

Typically, each HA server has its own IP address or host name. If you place the servers on the same
subnet, they can share the same IP using virtual IP addressing, which simplifies device configuration.

 

Once HA is set up, you should avoid changing the IP addresses or host names of the HA servers, as this will break the HA setup (see “Resetting the Server IP Address or Host Name” in Related Topics).

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

File and Database Synchronization

Whenever the HA configuration determines that there is a change on the primary server, it synchronizes
this change with the secondary server. These changes are of two types:

1. Database: These include database updates related to configuration, performance and monitoring
data.

2. File: These include changes to configuration files.
Oracle Recovery Manager (RMAN) running on both servers creates the active and standby databases and
synchronizes the databases when there are changes, with the help of Oracle Data Guard Broker running
on the primary server.

File changes are synchronized using the HTTPS protocol. File synchronization is done either in:

• Batch: This category includes files that are not updated frequently (such as license files). These files
are synchronized once every 500 seconds.

Near Real-Time: Files that are updated frequently fall under this category. These files are
synchronized once every 11 seconds.

By default, the HA framework is configured to copy all the required configuration data, including:
• Report configurations
• Configuration Templates
• TFTP-root
• Administration settings
• Licensing files

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

If there is a firewall configured between the primary and the secondary servers, ensure that the
firewall permits incoming and outgoing TCP/UDP on the following ports:

 

– 8082: Used by the Health Monitor process to exchange heartbeat messages
– 1522: Used by Oracle to synchronize data

 

• If you plan on using Operations Center with an HA implementation of Prime Infrastructure: Ensure
that all of your HA-enabled Prime Infrastructure servers (both primary and secondary) have fully
resolved host names.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.