on 06-17-2014 2:33 PM
Dear Experts,
I am getting an issue in Engine status, messages are in Tobedelivered State, Engine status --> backlog Monitoring also shows so many messages.
After some time those are going, but here when i am checking Addition Data--> Threads those are fully utilized 15/15.
We maintained Thread count is 250 and We are assigned worker threads for asyn and sync 15,15,12,12.
suggest me how to resolve this issue.
Regards,
Kiran Polani
Dear,
Some times we are getting so many queues in engine backlogs, but is going automatically. thread count is 300 only but as per those sap note that is not mentioned 350 to more. how much value i can increase.
Please suggest what can i do for this.
Regards,
Kiran Polani
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Dear all,
Thanks for your support, i found this error in SAP Idoc level and those are checked with processing inorder check box in SAP side. that's why Queues are going slowly one by one.
After unchecking that check box queues are going smoothly and increased worker threads for fastning Queues and given priority for that interface as per SAP Labs Suggestion.
Regards,
Kiran Polani
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hello Kiran,
Check the Thread's for each adapter, which i think you would see in technical details.
See the scenario for which you have high volume of messages (for e.g. - File Adapter has high volume) you need to increase that thread count. Also when you are facing with such slow processing issue, kindly take snapshot and check if the Adapter threads are available or not. Usually these would be fully consumed.
Regards,
Hiren
Hi Sri,
what is SAP PI version and do you have any third party components deployed in PI system like seeburger etc..?
how many java nodes in the PI system? what is the max heap memory allocated to Jlaunch processes? check for any memory leaks.
how many threads allocated to the queue ?
File_http://sap.com/xi/XI/SystemRecv
Cheers
Hidayat
Dear Ansari,
We are currently in sap pi 7.0, we are using only AtoA applications. like jdbc,rfc,soap,proxies. java nodes are one.
Yes | 0 | 0 / 0 | 15 | |
File_http://sap.com/xi/XI/SystemCall | Yes | 0 | 0 / 0 | 12 |
File_http://sap.com/xi/XI/SystemRecv | Yes | 0 | 0 / 0 | 15 |
File_http://sap.com/xi/XI/SystemRqst | Yes | 0 | 0 / 0 | 12 |
File_http://sap.com/xi/XI/SystemSend |
RFC_http://sap.com/xi/XI/SystemCall | Yes | 0 | 0 / 0 | 12 |
RFC_http://sap.com/xi/XI/SystemRecv | Yes | 0 | 0 / 0 | 15 |
RFC_http://sap.com/xi/XI/SystemRqst | Yes | 0 | 1 / 1 | 12 |
Yes | 0 | 0 / 0 | 15 |
SOAP_http://sap.com/xi/XI/SystemCall | Yes | 0 | 4 / 4 | 12 |
SOAP_http://sap.com/xi/XI/SystemRecv | Yes | 0 | 0 / 0 | 15 |
SOAP_http://sap.com/xi/XI/SystemRqst | Yes | 0 | 0 / 0 | 12 |
SOAP_http://sap.com/xi/XI/SystemSend | Yes | 0 | 0 / 0 | 15 |
JDBC_http://sap.com/xi/XI/SystemCall | Yes | 0 | 0 / 0 | 12 |
JDBC_http://sap.com/xi/XI/SystemRecv | Yes | 121 | 15 / 15 | 15 |
JDBC_http://sap.com/xi/XI/SystemRqst | Yes | 0 | 0 / 0 | 12 |
JDBC_http://sap.com/xi/XI/SystemSend | Yes | 0 | 0 / 0 | 15 |
1) we are allocating thread count is 250, what is the maximum value we can increase threadcount?
2) worker threads are assigend for system call is 12, systemRqst is 12 and system Recv 15 and SystemSend is 15. is these values are ok. or we can change.?
3) from which value we can change and can we change the value adapter wise.?
We are facing issue for soap and jdbc adapters those are using all resource from given.
Please suggest me what we can do for this.?
Regards,
Kiran polani
Hello Kiran,
Kindly check the below notes and you would have to use Poolwaiting time parameter.
Note 937159 – To be used for the stating our point of failure and increasing ThreadCount.
Note 821267 and 1473299– To be used for suggesting poolWaitingtime parameter.
This should fix your problem.
Regards,
Hiren
Dear,
As per that doc those parameter are set with proper values. but issue is not resolved. when load is come we are getting issue in XI.
Please advise.
Regards,
Kiran Polani
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Dear Harish,
It is not coming to all interface few particular business systems(3,4 BS at receiver end).
Regards,
Kiran Polani
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi Kiran,
If it is happening for SOAP receiver then you need to identify which pipe line step is taking more time.
Most of the cases performance issue is occure in transformation step (mapping), please check if you have any lookup (RFC/JDBC) in mapping. or because of large message size.
Please check the below document
blog
regards,
Harish
Dear Team,
I checked in Pipeline steps, we are getting delay in DB_ENTRY_QUEUING taking longer time to execute for few messages and for few DB_SPLITTER_QUEUING.
Please find the below screenshots for reference.
screen 1:
Screen 2:
Please suggest me what we can do for this.
We are maintained
In J2EE Visual Administrator --> messaging.connections
Messaging System parameters :
Send.maxConsumers=15
Recv.maxConsumers=15
Call.maxConsumers=12
Rqst.maxConsumers=12
is there any change required. Please suggest me.
Regards,
Kiran Polani
Dear Harish,
We unchecked in we20 "Queue Processing" in sap R/3 end for Idoc and unchecked in "Maintain Order at Runtime" in inteface Determination in XI level for avoiding EOIO format. Now those are going EO only with in XBT* Queues. But some times queues are going slowly and at a time in smq2 2000 messages are coming to XI end from R/3 Server, those signals going with 4, 5mins, but why suddently singals are coming to XI End & in Engine status Backlogs also we are able to see 1000 queues those are going but taking some time.
Regards,
Kiran Polani
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Dear,
All interfaces are getting this issue, those are Idoc to soap, Proxy to jdb, jdbc to proxy. It is not stucking long time but "Processing Backlog (without errors)" each Business service/system having 500 to 1200 messages. it is going slowly.
Please suggest..
Regards,
Kiran Polani
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi Kiran,
Message ToBeDeliver means that the message is sitting in the queue waiting to be processed.
In most situations, when this happens and stays there for a long time is because 1st message is the queue is being held up by an error.
BTW, what is your scenario ? what is your receiver system ?
Thanks
Hari.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
84 | |
24 | |
11 | |
9 | |
7 | |
6 | |
5 | |
5 | |
5 | |
4 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.