cancel
Showing results for 
Search instead for 
Did you mean: 

Performance issue and Error in ICM in case of a lot of messages

Former Member
0 Kudos

I am run performance testing for the following scenario.

csv files with ~1000 lines -> XML -> BPM - 1:N mapping - for each block- validation mapping - swith (choise JDBC or files System) send message.

My BPM has beed working for <b>2 hour</b>

The most messages are sent correnctly but

for some messages I see

<SAP:Code area="INTERNAL">CLIENT_RECEIVE_FAILED</SAP:Code>

<SAP:P1>405</SAP:P1>

<SAP:P2>ICM_HTTP_INTERNAL_ERROR</SAP:P2>

Message content is not big

for some

<SAP:AdditionalText>if_http_client receive http_communication_failure</SAP:AdditionalText>

<SAP:ApplicationFaultMessage namespace="" />

<SAP:Stack>An error occurred when refreshing the XI runtime cache</SAP:Stack>

Could you help me why this errors appear?

Message was edited by: Denis Ivanov

Accepted Solutions (0)

Answers (3)

Answers (3)

Former Member
0 Kudos

Hey Denis,

You can also search for SAP Notes using the transaction SNOTE

Regards,

Sushumna

MichalKrawczyk
Active Contributor
0 Kudos

hi Denis,

have a look at this note: 906435

it talks exactly about your problem I believe

Regards,

michal

-


<a href="/people/michal.krawczyk2/blog/2005/06/28/xipi-faq-frequently-asked-questions">XI FAQ - Frequently Asked Questions</a>

Former Member
0 Kudos

906435 location?

how can i find it ?

Former Member
0 Kudos

Hi,

Most of the time this error happens because of server overload.

Please can you check on the weblog for maintaing your system:

/people/sap.user72/blog/2005/11/29/xi-how-to-re-process-failed-xi-messages-automatically

Also, check if there are any queues blocked in transaction SMQ1 and SMQ2.

Regards

Vijaya

bhavesh_kantilal
Active Contributor
0 Kudos

Hi,

https://websmp202.sap-ag.de/notes

You need a Service market place id for this though.

Regards,

Bhavesh

udo_martens
Active Contributor
0 Kudos

Hi Dennis,

you have a problem with a http destination, may be not enough work processes. If you execute the test again, plz have a look

- to SM50: Are there enough dialog processes)

- to ST06 /detail analysis: Enough CPU power and enough memory?

- to ST22: Any short dumps?

- to SM21: Any critical system informations?

- to SMQ1/SMQ2: How many message are queued? How many errors?

- to SM37: Has Job with program RSQIWKEX restarted failed messages?

- SXMB_MONI: Are the failed messages restartable?

2 hours is defintivskaja to long. BPM is a performance bottleneck, espially if you have loops.

Regards,

Udo