cancel
Showing results for 
Search instead for 
Did you mean: 

Inbound Idocs: Queue Processing (non-stop queue?)

former_member182004
Contributor
0 Kudos

Hi everyone!

I have an issue with a File to Idoc interface. The customer would want the IDocs arriving to R3 to be in the same order they were sent from the legacy. From the File Adapter there isn't any problem using the EOIO QoS, and in the IDoc Receiver Adapter I checked the "Queue Processing".

There is something in the R3 side that the customer doesn't like much and it's about the stop of queue when an IDoc gets in error status: all the subsequent IDocs stay in status 75 ("IDoc received via qRFC").

I can change the status of the wrong IDoc and then release the stopped queue but: Is there a way to process serialized inbound IDocs avoiding the stop of the queue if an error occurs?

I'm using PI 7.0 SP30.

Thank you in advance!

Accepted Solutions (0)

Answers (1)

Answers (1)

Former Member
0 Kudos

First, this is not serialization, it's the queuing design EOIO... they are not the same.

So, since you activated EOIO in the Sender CC it's normal to have such behave (i.e. 75 status)...

Basically the serialization is not a good fit for these requirements.

Here is one solution:

1- Use dynamic queue name (via UDF) to create more parallelism since I dont think you have to queue all the IDOCs in one single queue.

2- Copy the IDOCs with 75 status to a new IDOCS with 64 (You must have very few)

3- Use RC1_IDOC_SET_STATUS to change the status of the IDOCS from 75 to something 73

4- There is  standard report that you have to schedule and restart the queues

5- Reprocess the IDOCs from steps 2

The enemy of EOIO design is the user locks in the target system...

Cheers,

Fouad

former_member182004
Contributor
0 Kudos

Hi Fouad, thanks for your response.

First of all, I need to respect the exact sequence so I can't use more than one queue. And about the 2nd point onward, If I send 1000 IDocs to R3 and 200 of them are incorrect hoy many time will take the complete processing of the IDocs?

Regards,

Juan.

Former Member
0 Kudos

You need to figure out why you are getting the 200 wrong idocs and try to fix the issue at the source at have just 20 next day.

The other solution is to go custom where PI does just the file transfer and a batch program will read each entry individually and posts the data in R3.

I faced the same challenge in a MES integration...

Cheers,

Fouad

former_member182004
Contributor
0 Kudos

Hi Fouad, thanks for your response.

I think that if you just use PI for moving the file and then use a batch process to enter the lines to R3 it won't be very practical to involve PI.
The idea of this post is about fixing this situation only with PI/R3 (not thinking in source legacies).

Thanks again.

Regads,

  Juan.

Former Member
0 Kudos

I listed the possible options,

Do you mind you post later what the design you adopted?

Cheers

Fouad

former_member182004
Contributor
0 Kudos

Hi Fouad,

The customer still didn't make a choice about the pros and cons of each possible solution so I'll tell you when it's implemented.

Regards,

Juan.

former_member182004
Contributor
0 Kudos

The customer preferred not to modify the interface. Thread closed.

Regards,

Juan.