Set specific health probes - Perceptive Content Server Install Guide - Foundation 24.1 - Foundation 24.1 - Ready - Perceptive Content - external - Perceptive-Content/Perceptive-Content-Server-Install-Guide/Foundation-24.1/Perceptive-Content-Server-Install-Guide/Install-Perceptive-Content/Assemble-and-configure-a-server-farm-for-Perceptive-Content-Server/Set-up-server-health-monitoring/Set-specific-health-probes - 2024-04-02

Perceptive Content Server Install Guide

Platform
Perceptive Content
Product
Perceptive Content Server Install Guide
Release
Foundation 24.1
License

To set a specific health probe for Perceptive Content Server, complete the following steps.

  1. Open the Cisco ACE configuration tool.
  2. Click the Config tab.
  3. In the Load Balancing menu, select Health Monitoring.
  4. In the Probe name field, provide a name for the probe.
  5. In the Health probe type field, select TCP for the type.
  6. In the Port field, enter 6000. This is the port that Perceptive Content Server listens on, as defined in the inserver.ini file.
  7. In the Send Data field, enter the following information.
    000000011SERVER_PING000000008END_MARK000000013CLOSE_SESSION
  8. In the Expect Regular Expression field, under the More Settings header, enter the following information.
    .*SUCCESS.*
    
    Note: The health probe ensures that SUCCESS appears in the returned string. A SUCCESS response indicates the server is connected to the database and accepting logins. If .* tokens are not valid on your load balancer, adjust the configured receive string. Load balancers that perform a 'contains' comparison with the receive string do not require the .* token.
  9. In the Probe interval count field, enter a probe interval count. This is the time interval between sending probes during a health check. The recommended value is 15 seconds.
  10. In the Pass detect (failed probe) interval field, enter a pass detect count. This is the time interval between sending probes during a health check when the server is in a known bad state. The recommended value is 30 seconds.
  11. In the Fail detect count field, enter a fail-detail count. This is the consecutive number of times a probe must fail before the server is marked as failed. The recommended count is two times.
    If you set the Probe interval count to 15 seconds, the load balancer stops sending new connections to the server within 30 seconds of it going down.
  12. In the Pass detect count field, enter a pass detect count. This is the number of successful responses a probe must produce before the server is marked healthy. The recommended count is two times.
    If you set the Pass detect (failed probe) interval to 30 seconds, the load balancer starts to send new connections to the server within one minute of it coming back up.
  13. In the Receive timeout for a response count field, enter a receive timeout for a response count. This is the amount of time that a server has to return a response during a probe. If it does not return a response within the set time, the probe fails. The recommended value is five seconds.