Hopefully someone has a thought about this. I have a single fex that has a few other -
INCLUDES. I was under the impression that a single fex would tie to a single agent, but I have long running report fex, due to the SQL, but I see 3 agents spawn (not all together but as needed) but I see they are making DBMS calls.
The server doesn’t have any limitations on the agent itself, so the agent has full run of the house.

I had the understanding that a –INCLUDE would use the same agent, but a –EXEC would spawn a different agent. There are no –EXEC’s in the program.

Is it possible for the database to multi-thread the SQL on its side, but spawn agents on the Webfocus side?

This message has been edited. Last edited by: Kerry,

I think -MRNOEDIT INCLUDE spawn only one agent. As I understand it, the MRNOEDIT pulls in the INCLUDEd code at run time whereas a straight INCLUDE spawns out.

I could be wrong.

- ABT

It goes top-down, and these agents are DBMS calls, so if the first SQL is taking "X" amount of time to return, I don't see how it can move on to do further processing to spawn additional agents.

Thanks for the quick response.

I have seen some examples of agents with very long running RDBMS calls hit some sort of time limit and spawn a second agent.

I thought that is was fixed.

How many records is being read/retrieved.

How long are the agents running ?

Depending on selection, no more then 2,000 records are being brought back to Webfocus. Just eye balling it, roughly about 8 minutes or more.

We are working on the performance of the SQL call to reduce as much as possible, but the concern is concurrency usage with the way my server is configured with agent max and queing turned on.

Need to understand what my user to agent ratio is or in this case fex to agent ratio is.

This message has been edited. Last edited by: MattC,

Worked with IBI on this issue and we have a possible solution. It's relates to using Weblogic as a Web Server. There is a setting on the Weblogic side called

WLIOTimeoutSecs (new name for HungServerRecoverSecs)

- Defines the amount of time the plug-in waits for a response to a request from WebLogic Server.

Basically as the RDBMS call is waiting for a response as it surpasses this wait time a new agent spawns. Working with our server folks to make a change to increase this, so I will update this post if this does solve the issue.