|
|
A server's availability is measured by the percentage of time, it is providing service, in the total time since it is deployed. In general, the servers are expected to be available for 24 hours a day, 7 days a week, and 365 days a year. However, a part or all of the server may be down, during a planned or unplanned downtime. Now, here comes the need for the High Availability, which provides fault tolerance and improved scalability.
IT360 supports the HA environment, with the Primary and Standby servers installed on different machines.The below points will explain the working of the HA setup, in detail;
The High Availability set up requires a couple of IT360 installations; one is the Primary (Central / Probe) server and the other is the Standby (Central / Probe) server. Each one has its own MySQL backend database.
Initially, the Primary (Central / Probe) server will be up, running and providing service. On the other hand, the Standby (Central / Probe) server will be started in hot standby mode, where it keeps on monitoring the Primary server, without providing any service. However, it will be in a ready state to start itself as Primary, when the actual Primary server goes down.
Both the servers write their heart beat counts in the database. Hence, the Primary monitors the heart beat count, written by Standby, and vice versa, periodically.
When the heart beat count of the Primary is found to be not incremented, the Standby server will consider that the Primary has gone down. Hence, it will try to start itself as the Primary server. Before doing so, it will perform an additional test to ensure that it is able to reach the Primary, using HTTP connection (HTTPS / HTTP port).
Similarly, when the heart beat of the Standby is found to be not incremented, the Primary will consider that the Standby has gone down. As a result, it will de-register the Standby from its scope.
Click on the links below to view the explicit procedures, for configuring HA for the Enterprise Edition (MySQL) - Central & Probe;
|
|