Active healthchecks are disabled by default and can be enabled for a server pool by binding a health monitor to the Group through the LBRule object. This represents active health monitoring over TCP. Active healthchecks are initiated periodically, at a configurable interval, to each member of the Group. Only if a healthcheck fails consecutively for a specified number of times (fall_count) to a member will the member status be marked DOWN. Once a member is DOWN, a specified number of consecutive successful healthchecks (rise_count) will bring the member back to UP state. After a healthcheck is initiated, if it does not complete within a certain period, then also the healthcheck is considered to be unsuccessful. Completing a healthcheck within timeout means establishing a connection (TCP or SSL), if applicable, sending the request and receiving the response, all within the configured timeout.


This class contains all of the following classes.

Was this page helpful?