As I reported earlier, we start now only one EntireX RPC Server job, which contains 5 tasks (NTASKS=5) and use TCP/IP for communication. Before, we started 5 Server jobs using SVC.

It was no problem to terminate the Server job by issuing the TERMINATE command through user exit (API) USR2071N. Now, with 1 job and 5 tasks, we have to terminate five times the server. Even one cancel command (CO) for the job won’t work. We have to do it five times.

Before the change, when the Broker ended, the Server also ended automatically. Now, the Server job is up and running but nobody can use the server.

Any ideas?

Dieter Storr

This sounds like you are not using ADALNKR, which is a prerequisite for RPC servers with NTASKS > 1, see

[url]http://techcommunity.softwareag.com/ecosystem/documentation/natural/nat424mf/rpc/start.htm#start_Considerations_for_Natural_RPC_Servers_with_Replicates[/url]

Thanks, Wolfgang.

I will discuss it with the system folks.

Dieter