I am using Biztalk server 2006.
I have a SendPortGroup with 3 SendPorts. I use filter in SendPortGroup such that if there is a message fulfill the filtering criteria, 3 SendPorts should send this message to the corresponding outbound folders.
Today I found that only 1 SendPort send the messages, the other 2 do not.
After restarting the Send and Receive host instance, it resumes normal, 3 SendPorts can send message.
During the abnormal stage, there is no error event captured in event log.
Do you have any idea on that abnormal situation?
Never noticed this kind of behaviour.
I would recommend enabling tracking on the Ports and next time before you restart the hosts check the Eventlog for any unusual error messages and investigate the issue.
Regards,NISHIL MCT,MCTS | Freelance BizTalk Consultant.biztalkconnect.blogspot.com
I hope you have installed all the latest Service packs for BizTalk 2006 R2. The support from Microsoft on BizTalk 2006 R2 has expired on 7/12/2011. I would recommend you to plan moving to the latest version 2013R2 which will be released soon asap.
In Biztalk running service instance, I found the 2 sendport service instance and those not-sent messages can be found the message tab and the message status is “In process”, but in fact the process should be completed a few hours ago.
Do you have any idea on that?
it is found that memory used by one Biztalk host instance is kept increasing, without releasing. Any idea why the host instance memory is kept increasing, without releasing? thx
1. How much memory is this host using?
2. Have you disabled tracking on this host?
3. Are you on a 32 Bit BizTalk configuration?
4. Are all the BizTalk SQL jobs running fine?
5. Do you have dedicated tracking host?
6. Do you a seperate host per Adapter type?
7. Is it a heavy message processing scenario when this issue occurs(large messages/many in concurrent messages)?
You are clearly having a memory problem...where in under heavy load this Sendport doesnot have enough resources to complete the operation.
There are many reasons for memory issues.
1. the largest memory used by one host instance is 2.9GB.
2. host tracking has been disabled
3. 64 bit
4. running fine
7. around 10 of concurrent messages
If look at the perfmon, i got the PID of the host instance, but how I can know which PID refer to which host instance?