TargetLoadBalancing Feature - KeyInfo

[asribalaji]

Quick information on the TargetLoadBalancing feature usage:

1. Management API to check to a target is reachable or not. Configure the target info(ip & port) as a target server under an environment & use the following MGMT api to check whether the target(connection) is reachable or not.

POST v1/organizations/{organization}/environments/{environment}/targetservers/{targetserver}/?action=test 

2. LB configuration to mark a target as down & take it out of the target list serving the runtime requests.

LoadBalancer configuration <MaxFailures> (as shown below) used to determine the no of the times a target should be checked for availability before marking it as down & take it out of the available target list.

<TargetEndpoint name="default">
  <HTTPTargetConnection>
      <LoadBalancer>
        <Algorithm>RoundRobin</Algorithm>
        <Server name="target1" />
           <MaxFailures>5</MaxFailures>
        <Server name="target2" />
      </LoadBalancer>
      <Path>/test</Path>
  </HTTPTargetConnection>
</TargetEndpoint>

3. Monitoring support in TargetLoadBalancing

With HealthMonitoring configured support (TCP or HTTP based), the target backup/availability check can be achieved after it has been marked as down.

<TCPMonitor> & <HTTPMonitor> configurations can be used to check for the target availability on a scheduled interval & add it back into the available target list.

Also to handle target (reachable)failures, In the <LoadBalancer> configuration, one of the target servers can be configured as a backup/fallback server (highly available) i.e in case none of configured target server is available in the target list, the fallback server will be always in place to serve the requests.

<TargetEndpoint name="default">
  <HTTPTargetConnection>
  <HealthMonitor>
    <IsEnabled>true</IsEnabled>
    <IntervalInSec>5</IntervalInSec>
    <TCPMonitor>
      <ConnectTimeoutInSec>10</ConnectTimeoutInSec>
      <Port>80</Port>
    </TCPMonitor>
  </HealthMonitor>
. . . 

More details on the loadbalancer(monitoring,fallback) configuration :

http://apigee.com/docs/api-services/content/load-balancing-across-backend-servers