Portal working with two nodes yet unavailable when one node is down

[Former Member]

Hello,

Our Portal is running with two nodes.

When we stop one server the whole portal is not available.

Where can I verify that the portal is working as a cluster and what could be the reason for that?

Regards,

Roy.

[mike_chen]

Hello Roy,

So you are working in a single server environment. However, when you shutdown one server node, the other should still work. How do you shutdown the server node? Use MMC, jcmon or simply kill process? Please refer to the admin document and make sure that you did it in a proper way. Also, please check the /JC<instance_ID>/work folder, pay more attention to the std_server<node_number>.out, and other updated files. To see if there are any error messages there. Also, please have a look at the defaultTraces, to see whether there are any errors.

BR/Mike

[mike_chen]

Hello Roy,

What does the "two servers" here mean? Does it mean one physical server with two server nodes? Or does it means two separate dialog instances? Also, how is your portal used? Are there any loadbalancer or reverse proxy involved? What's the version of your portal? If it is new enough (later than SP14), please try /irj/servlet/prt/portal/prtroot/com.sap.netweaver.bc.rf.explorer.RFExplorer, with this tool, it will tell you how many server nodes currently are working, and on which server nodes the portal is currently running.

Best Regards,

Mike

[Former Member]

Hello Mike,

Thank you for your response, I apologize for not being clear enough.

When I say "two servers" I mean that the portal is running on one machine with two server nodes. Currently we are not using any dialog instance (By the way, what does it use for?) I am using SP14.

Perhaps there is a dispatcher configuration perhaps that tells the portal to which server to forward the requests...?