SXMSPMAST is huge and growing every day!!!!

[former_member190996]

Good day,

We have quite a few ABAP proxies which are synchronous messages and we the LOGGING_SYNC is 1. For different reasons we are storing the sync messages on ECC. But now from SXMB_MONI, if I select any message to display it's taking forever and timing out. Then I have started looking at the SXMSPMAST, and to my shock there are aroung 27millions+ records in the table.

So, I started looking at the messages with INIT/ARCH/DEL status. But 99.9% of the messages are in INIT status:

We have delete job running every day, and when I check it's log, it says records deleted, but that's not the case. The no. of records in SXMSPMAST are increasing everyday, and I'm very much concerned about the performance impact due to this:

Delete job log:

Appreciate any kind of suggestions to overcome this.

Best Regards,

SB

[weberpat · Accepted solution]

Hello Sai,
First of all, I'm more of a solution consultant rather than an expert on IE administration but for all I know, the cleanup jobs scheduled via SXMB_ADM would only pick up messages that have their interface action set to DEL or ARCH. The INIT status is set when the message enters the Integration Engine but then changed immediately to either ARCH (if the interface is configured for archiving) or DEL (if no archiving has been defined). The logic according to which this change happens is coded in method CHECK_INTERFACE of class CL_XMS_PERSIST, so you may want to have a look and see if there is any condition that could prevent the status from being changed and is applicable in your scenario.

Probably less than you have hoped for but I hope it can set you on the right path to find out what's going on.

Regards,
Patrick

[former_member190996 · Accepted solution]

Hi Patrick,

Appreciate your response, will try to debug the programs that are being used to archive/delete the messages.

Regards,

Sai

[former_member190996]

Any suggestions?? @vadim.klimov??