SMQ2 - Status: RUNNING forever?

[Former Member]

Team,

We are presently processing a lot of messages – all asyn.

In SMQ2, we can see all (XBTI) inbound queue messages and (XBTO) outbound queues messages.

At some point we get CCMS errors on queues saying that the queue are blocked with the status RUNNING. According to the documentation, a RUNNING status mean that “The first LUW of this queue is currently being processed.” … but the queue can stay with this status for ever .. unless we save the LUW (to SMQ3), and/or - unregistrer/registrer queue - then the queue unblock it self and at the end of the run, we restore the LUW back and it pass…

Is there anyone who can explain the behavior ?

Something we get the status “READY”, again according to the documentation, this status (is temporary) mean that there is a lock happening behind the scene. We can unlock the queue and again, the queue get process.

What is the difference between un registering /registering queues and unlock/locking them ?

Thank a bunch in advance

[VijayKonam]

Hi,

I had faced this problem badly and researched what all I could.

The Q status is Executing for ever, then there could be a dead lock inside IE, in processing the message and the system went into infinite loop. There is no otherway other than stopping(Lock immediately) the Q, and then unlocking it.

If any Q is under sysfail condition, there must be something wrong with the message and please check ST22 for any short dumps to analysis.

When you just delete a single LUW, you need not have to register the Q. But when you delete a Q itself, it is required to re-registered in SMQR. Unless, those Qs will not be created at run time by the system.

Be carreful that, you should not delete the Q unless u do not want the other waiting messages in side the Q should not be processed.

VJ

[Former Member]

Hi VJ,

Thank for you response

Could it be because we are running out of work process ?

Also, the configuration of the generic queue (XBTI*) in (SMQR) are (max runtime 60 seconds) ... So if QIN goes over 60 second, the documentation say that it stop processing the current LUW and does not pickup the next LUW in line ..

Thank again !

A+

[Former Member]

Hi all,

Is there anyone who ever had transactions in SMQ2 that has the status RUNNING and stay in that status for ever ? The only way to restart the queue is either by registering it again and-or by unlocking the queue.

We checked in our XI + in each and every backend system (business system) and there is a enough work process ...

What else can cause that ... help help help

Michel Osborne

[Former Member]

Hello,

I had similar problem on PI running on MSCS. I solved it by deleting the redundant Enqueue process on Abap stack (rz10), and defining two enqueue processes “enserver.exe” on java stack – one for ASCS, and one for SCS. You can find more information in installation guide for your XI type.

Janos

[Former Member]

Hi Janos,

This is exactly how we are setup. The difference is that we are using MCServiceGuard (HP) ...

Our QA & Prod are setup exactly the same ... We are not able to reproduce the behavior in QA, but it happen in Prod on a regular basis

A+