BIG-IP Monitoring of Front-End Web Servers and Server Health (Windows SharePoint Services 2.0)

  • Article

TCP/IP PING was enabled on each front-end Web server in the load balancing pool. To make sure that each front-end Web server responds to TCP/IP PING and Windows SharePoint Services HTTP requests properly, the Internet Platform and Operations group defined a test rule by using the Extended Content Verification (ECV) Monitor supplied by the BIG-IP controllers. Appropriate settings for Send String, Receive Rule, Username, and Password were entered to test the Windows SharePoint Services site response from each included front-end Web server. Send String was defined as accessing the default.aspx page of a test SharePoint site in the server farm. The response of the request should include the substring keyword "Home." Therefore, the Receive Rule is defined as "Home."

For the test rule, the interval is defined as 15 and the timeout is set as 46. The BIG-IP controllers will issue a request every 15 seconds to each front-end Web server. If no appropriate response is sent back in 46 seconds (three retrials), BIG-IP will interpret that this front-end Web server has certain application issues and can not appropriately render Windows SharePoint Services requests. It will automatically remove the front-end Web server from the load balancing pool in this situation. After the issue is resolved, the BIG-IP controllers receive appropriate Windows SharePoint Services responses and add the front-end Web server back to the load balancing pool. The interval can be shorter or longer depending on the observed traffic load and system behavior. The shorter interval may not allow the front-end Web servers to respond quickly enough during a heavy load situation.

The TCP/IP PING and test rule provide a monitoring solution to the Windows SharePoint Services server farm. If a front-end Web server encounters any critical application issues, the front-end Web server will be automatically removed from the load balancing pool. MOM will automatically report the issue to administrators. End users would not experience rendering failures from front-end Web servers with problems. This design and implementation help ensure that the Windows SharePoint Services server farm provides high availability and reliability.