Foglight Evolve 9.3 - High Availability Field Guide

Running on Oracle RAC

In this configuration, the Management Server was installed on two distinct physical hosts, one acting as HA mode primary, and the other as secondary. This setup was run over an Oracle® RAC cluster to provide HA at the database layer. In addition, an HTTP proxy was configured to provide end users with a single URL for access to the Foglight® browser interface.

NOTE: For more information about Oracle RAC, see: http://www.oracle.com/technetwork/database/clustering/overview/index.html.

NOTE: The previous procedure outlines only one method of configuration. Other methods include: using the SQL scripts included with Foglight to do the first-time configuration, and further configuration of the oracle-ds.xml file. These methods are possible, but have not been tested.

Running behind an HTTP proxy

One of the most common configurations is running the Management Server in HA mode behind a proxy so that end users and the agent configuration only need one URL for connection to the browser interface.

In the following example, a third-party software program called balance is used as a proxy.

NOTE: On failover to another member of the cluster, the Foglight browser interface session will restart.

Running in a cluster

This section describes one scenario for running the Foglight® Management Server in HA mode with clustering at the application layer. In particular, the focus is on operating system level clustering used in commercial systems such as Oracle®, Microsoft®, and VeritasTM clusters.

This scenario uses the Linux® HA Project’s heartbeat software for SuSE ES 10.1. The heartbeat package includes a virtual IP resource called IPaddr2.

NOTE: The agent has no mechanism to tie it to a particular interface, and it stops sending data to the Management Server at failover until restarted.

For more information about the Linux HA Project and heartbeat, see http://linux-ha.org.

Tuning connection issues in HA implementations

In some cases, the communication between servers in the cluster is not reliable. Sometimes unwanted behaviors occur, such as:

These issues may be attributable to JGroup (the underlying communication package that the Foglight Management Server uses for its HA implementation) and to the fact that a Foglight HA implementation uses UDP for communication by default and UDP is by nature an unreliable protocol.