SAS WebApp | Horizontal Cluster -- 404 after start the second mid tier

[MariaD]

Hi folks,

We have the following configuration: 1 compute server (meta and mid tier installed) + 2 horizontal cluster for mid-tier. All servers are Linux and 9.4M7, we user mostly SAS VA for WebApp.

The plan used to install and configuration was sent by Technical Support. We have the following situation:

If we start compute and web 1 server, everything works fine. SAS VA works perfectly fine. After this verification, if we start web 2 (starting occurs fine without any error), when we tried to open SAS VA, a 404 error message appears. The only solution is stop web 2.

But if we start compute node and web 2 server, everything works fine too. SAS VA works perfectly fine. After this verification, if we start web 1 (starting occurs fine without any error), when we tried to open SAS VA, a 404 error message appears. In this case, the only solution is stop web 1. 

Any idea why we can have only one of the web node started?

Regards,

[dpage]

Can you confirm if access to the VA app directly on both the backends works without a 404 (I'm curious if the 404 is coming from the upstream server, or the reverse proxy directly)

You can get the info from LevX/Web/WebServer/conf/sas.conf - Check ProxyPass for the /SASVisualAnalytics entry, see what balancer it refers to, go to that section and you'll 

see listed BalancerMembers for the backend, with an explicit port like 8080 or 8180

It will redirect you to the logon, you can't use the app directly this way but you can at least confirm the backend isn't impacted. 

I would open a Tech Support track and they can help, I would also review and attach the logs from the webserver, and relevant web app servers. 

[MariaD]

Thanks, I'll open a ticket and let you know how's end.