| Author |
Message
|
| mqxplorer |
 Posted: Mon Aug 18, 2014 8:13 pm Post subject: Issue with Invalid SOAP Reply Identifier |
 |
|
Master
Joined: 22 Jun 2009
Posts: 206
|
Hi,
By looking at the subject line, most of the people might not even look at my post as it is a very common basic problem in WMB.
However, I have a weird situation where I have below exception in my logs...
| Quote: |
| BIP3742E: An attempt was made to use a SOAP message ReplyIdentifier after it had already been used |
This is happening only when I do volume test. Let me explain my interface.
I have a DataPower interface with input queue where another WMB flow writes messages to and DP interface picks up the messages and calls another WMB web service flow and based on the reply to queue info that was sent in the input request, the response from the back end WMB flow will eb sent to the appropriate queue. This works prefectky fine with fewer no of requests sent across to the input queue configured for the DP interface. However, when we do volume test in Perf environment (in this case the volume is just 8000 messages), all the responses do not come back to the resp queue and somtime there is adelay for the responses. The WMB web service flow that receives the requests from DP has some weird issue which I am not able to figure out after spedning a lot of time. In my wmb logs, I see the error message as below...
| Quote: |
| BIP3742E: An attempt was made to use a SOAP message ReplyIdentifier after it had already been used |
and at the bottom I see the below error message..
| Quote: |
| BIP3761E: The input node ABC.SOAP Input' made an attempt to send a fault, but a reply had already been sent. |
Based on my analysis, the only conclusion I could come to was... some response like a Timeout or some wried resp is sent back to DP from WMB flow and when the WMB flow is trying to send the actual resp/fault response back to DP, it says that ReplyIdentifier is alrready used. I might look like a complete dumb guy when you read what I wrote but I am totally lost and trying really hard to figure out how come a ReplyIdentifier is already used.
I see some similar but not same issue as per the below link...
http://www-01.ibm.com/support/docview.wss?uid=swg1IT00268
However, this is mentioned as an issue in V9. I am Version on 7.0.0.5.
Please help me understand the root cause for this.
Thanks
mqxplorer |
|
| Back to top |
|
 |
| McueMart |
| Posted: Mon Aug 18, 2014 10:08 pm Post subject: |
|
|
Chevalier
Joined: 29 Nov 2011
Posts: 490
Location: UK...somewhere
|
A fair bet would be that v7 also exhibited the problem, but IBM decided only to fix it in v9 (for some reason or other).
I would open a PMR to have this clarified and hopefully have them provide you with a v7 APAR as it's still in support (for the next year at least!). |
|
| Back to top |
|
|
| mqxplorer |
| Posted: Tue Aug 19, 2014 5:10 am Post subject: |
|
|
Master
Joined: 22 Jun 2009
Posts: 206
|
| McueMart wrote: |
A fair bet would be that v7 also exhibited the problem, but IBM decided only to fix it in v9 (for some reason or other).
I would open a PMR to have this clarified and hopefully have them provide you with a v7 APAR as it's still in support (for the next year at least!). |
As per the problem description in http://www-01.ibm.com/support/docview.wss?uid=swg1IT00268 it says...
| Quote: |
Intermittent SOAP Reply Identifier exceptions reporting that
the reply handle is not valid might be seen under heavy message
load when the SOAPReply node is in a separate message flow to
the SOAPInput node.
|
However, in my case both SOAP Input and Reply nodes are in the same flow.
Would you still recommend opening a PMR?
Thanks
mqxplorer |
|
| Back to top |
|
|
| fjb_saper |
| Posted: Tue Aug 19, 2014 5:20 am Post subject: |
|
|
Grand High Poobah
Joined: 18 Nov 2003
Posts: 20756
Location: LI,NY
|
| mqxplorer wrote: |
However, in my case both SOAP Input and Reply nodes are in the same flow.
Would you still recommend opening a PMR?
Thanks
mqxplorer |
It might still be the same problem. There are a few different threads in play (specially with time-out management) and the synchronization of the access t the reply identifier might still solve your problem.
Now if you do get a response / failure that you want to send to the client be aware that if that response is served after expiry of the time-out interval, the time-out message will have used the reply identifier and the response you see is to be expected.
Have fun 
_________________
MQ & Broker admin |
|
| Back to top |
|
|
| ganesh |
| Posted: Thu Aug 21, 2014 11:40 am Post subject: |
|
|
Master
Joined: 18 Jul 2010
Posts: 294
|
Fix for BI3761E was supposed to be in 7.0.0.6 according to this article - http://www-01.ibm.com/support/docview.wss?rs=171&uid=swg1IC90615 but even after installing 7.0.0.6 we still see this error.
According to this article ->http://www-01.ibm.com/support/docview.wss?uid=swg1PI16711 the issue is present in all versions of 7,8 and 9 and the fix will be available in 8.0.0.5 and 9.0.0.3, nothing has been said about fix for version 7. |
|
| Back to top |
|
|
| smdavies99 |
| Posted: Thu Aug 21, 2014 10:53 pm Post subject: |
|
|
Jedi Council
Joined: 10 Feb 2003
Posts: 6076
Location: Somewhere over the Rainbow this side of Never-never land.
|
If this is really the fix you need then raise a PMR quoting this thread plus the fixPack announcements and IBM may be able to :-
1) Verify that this is indeed the fix and supply you with a fix
2) Say that the fix was in the latest V7 fixPack and suggest ways to modify your code in order to avoid the issue.
It might be that you have found another problem with the code that is not covered by the latest fix.
_________________
WMQ User since 1999
MQSI/WBI/WMB/'Thingy' User since 2002
Linux user since 1995
Every time you reinvent the wheel the more square it gets (anon). If in doubt think and investigate before you ask silly questions. |
|
| Back to top |
|
|
|
|
