To view the latest 11.3.x version, see  PATROL Agent 11.3.02 Open link .

Failover behavior

The PATROL Agent supports failover as described and illustrated in this section.

The following table describes the sequence of events that occur during a failover in a cluster environment and how the PATROL Agent behaves.

Note

Although the following table uses Windows cluster terminology, the agent behavior applies to both Windows and UNIX clusters.


PATROL Agent behavior in a cluster environment

SNo

Events

Description

Illustration

1

Request

A user requests that the cluster application be started. A request can occur at anytime, including startup.

Request, start up, and run a cluster application

2

Start up

The cluster management application runs the application on host A1. It starts the group which starts the cluster application and starts the PATROL Agent defined within the group. The agent reads the configuration file stored on a shared drive and starts up based on the file information.

Request, start up, and run a cluster application

3

Run

The cluster application performs services and writes data to the shared disk drive. The agent monitors the application and writes the history to the file specified by PATROL_HISTORY_<port>.

Request, start up, and run a cluster application

4

Service Interrupted

The cluster application is unable to access resources that it requires to run. The reason for this interruption can range from the failure of hardware or software to a forced failover by a system administrator.

Service interruption and shutdown of a cluster application

5

Shutdown

The cluster management software shuts down the cluster application and the PATROL Agent monitoring it on host A1.

Service interruption and shutdown of a cluster application

6

Failover

The cluster management software starts the group on B2, which starts the cluster application and the PATROL Agent.

Note
Local KM customizations stored on A1 are not available for the PATROL Agent on B2.

Failover and resumption of a cluster application

7

Resume

The cluster application performs services and writes data to the shared disk drive. The agent on host B2 monitors the application now running on B2 and writes history to the file specified by PATROL_HISTORY_<port>.

Failover and resumption of a cluster application

The following figures graphically depict how the PATROL Agent behaves during a failover in a cluster environment:

 Request, start up, and run a cluster application 

 


 Service interruption and shutdown of a cluster application 



 Failover and resumption of a cluster application 

 

Environment variables 

WindowsUNIX
PATROL_VIRTUALNAME_8888=PATROL_DiskPATROL_VIRTUALNAME_8888=PATROL_Disk
PATROL_HISTORY_8888=K:\historyPATROL_HISTORY_8888=/history
PATROL_CONFIG_8888=K:\config PATROL_CONFIG_8888=/config 
Was this page helpful? Yes No Submitting... Thank you

Comments