on 03-02-2009 8:48 AM
Sometime when we send a lot of messages to JDBC adapter, we get this exception in the JDBC communication channel:
Delivery of the message to the application using connection JDBC_http://sap.com/xi/XI/System failed, due to: com.sap.aii.af.ra.ms.api.RecoverableException: Channel has reached maximum concurrency (5,000 concurrent messages) and no free resource found within 5,000 milliseconds; increase the maximum concurrency level.
When I restart the XI, the problem solved temporarily, but after every several days,it came out again.
So I checked some parameter in VA, I found these two paramaters which have the value 5000:
MaxFOQSize Maximum size of pool with reusable objects 5000
MaxRQSize Maximum size of waiting object queue 5000
I wonder if these two parameters are the right parameters which I should change.For example, I tune the two paramters to 10000 both.
Any Sugguestion?Thanks a lot!
Hi,
Jdbc adapter communication channel in the advanced mode mark the parameter "Disconnect from Database After Processing Each Message".
Regards,
Prakasu
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi,
The load on the database will be very much high if you allow 5.000 concurrent messages, so is the error.
i recommend you to revise the scenario architecture.
Regards,
Swetha.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi,
If you want process n no of messages concurrently, you can set the Maximum Concurrency level=n in the Receiver Communication Channel For File/JDBC in PI 7.0 SP 12, .
By Default it is 1.
Also check,
Regards,
Swetha.
This would mean that channel has 5000 open database instances which is too large. Instead of thinking about further increasing it, it would be better to redesign if possible.
Regards,
Prateek
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi Prateek, this time it happened to file adater. In my scenario, I use XI to send the idoc DESADV02 from R3 to multiple file receivers. And I checked the successful message numbers from the adapter engine, it's already more than 10000 numbers. I don't know what kind of message occupied the database thread and did not release the resource.
Regrads,
Yitian
User | Count |
---|---|
87 | |
10 | |
10 | |
9 | |
7 | |
7 | |
6 | |
5 | |
4 | |
4 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.