cancel
Showing results for 
Search instead for 
Did you mean: 
Read only

Inbound queue Queue stuck problem

Former Member
0 Likes
3,272

Hi,

In the inbound queue(transaction - smq2), we found that lot of CSA* queues and R3A* queues were in stataus running. But the entries in the queue were not getting reduced even after 2 days. When double clicked I found old data with timestamp older than 2 days. The queues also has statustext "transaction executing" but nothing happened.

So to solve this problem in a short way,

1. Deregister the queue.

2. Manually execute one entry in the queue.

3. Register the queue back.

Now the queue started running properly. This is a short term solution and it needs an administrator to look daily if the queue are running or not. But we need to know why these queues get stuck and still have the status "running"

Looking forward from middleware experts why the queue gets struck and how to solve this problem .

kind regards

middleware developer

View Entire Topic
Former Member
0 Likes

Hi,

If the queue is in status running, this means that the queue-entry has started processing. If they stay like that, a number of things might have happened:

1. the process is still running: you can check processes in transaction SM50.

2. The process was aborted and no status was given back to the queue: check transaction st22 for short-dumps and transactions sm12 or sm13 for lock entries & update errors.

3. The process has ended, but the status update wasn't processed.

Either of these is a serious error and should be treated.

Search for Bdocs with an "intermediate" status: they are probably there waiting for the updates of the queue-entries (transaction SMW02).

Let me know how it turns out,

Michael.

Former Member
0 Likes

Hi Michael,

I checked all the 3 scenarios mentioned. None of them fits. There is an OSS note 758677, Let you know Once I apply the note,if this solves the problem or not. I think this is relevant.

thanks

middleware developer

Former Member
0 Likes

Geetha, without any problem you can add this note. The re-scheduling of your jobs can definately be you solution (it did for me at one of my projects). If this is not solving your solution check for the infinite loop mentioned and follow the notes, advise. Create an OSS message, Let me know what you find, Tiest.

Message was edited by: Tiest van Gool