Load Balancer Health Checks failing

[GillPaterson]

We have set up a highly available environment with Portal, Server and Geoevent and are using a Radware load balancer. The load balancer requires to do a health check of the individual machines within the group using the following URL.

https://Portal macine name/gisportal/portaladmin/healthCheck
https://AGS machine name:6443/arcgis/rest/info/healthCheck
https://GE machine name:6443/arcgis/rest/info/healthCheck

When the individual machines are checked manually through a browser, the status is returned as ok. When checked through the Radware lb, it fails. Has anyone used Radware before and had the same issues and resolved it please?

We have checked IP addresses and that the correct ports (7443 and 6443) are registered in the LB.

Thanks in advance.

[JonathanQuinn]

What is the response code of the health check? Are you going from LB -> Web adaptors -> Portals?

[GillPaterson]

We do have web adaptors in the mix and for Portal we seem to be using the WA address as it uses gisportal/portaladmin/healthCheck, but for the two AGS sites we are going straight through to the servers. If we were checking through the WA, the URLs would be:

- gissite/rest/info/healthCheck
- gisgesite/rest/info/healthCheck

I will try and find the exact error message from the LB team, they have only passed on that it fails.

[GillPaterson]

It turned out that the person who had entered the LB rules hadn't put a capital 'C' in healthCheck, and hadn't checked the logs (they claimed that there were no logs, they were proved wrong). All working fine now.