| Author | Message | 
		
		  | TheBigEasy | 
			  
				|  Posted: Tue Dec 06, 2005 8:34 am    Post subject: WMB v6 Broker Debug Problems |   |  | 
		
		  |  Apprentice
 
 
 Joined: 04 Jul 2005Posts: 33
 Location: London
 
 | 
			  
				| Hi 
 I have successfully been running a full WMB v6 proof of concept on a laptop:
 Configmgr using Cloudscape
 Broker using Db2 database
 
 The installation keeps falling apart following attempts to use debug. I have been able to successfully debug a flow several times. for more detail I added tracing - debugged another test message until completion and then stopped the tracing and terminated and removed the debug.
 
 Any attempt to start the broker results in the following error in the Event viewer and the cpu hits 100%. The toolkit can still bind to the ConFigMgr fine but no communication with the broker.
 
 I have deleted the broker and the db2 database and recreated them with the same name - problem remains the same.
 
 Any help to stop head scratching greatly appreciated.
 
 Error from Event viewer:
 
 
 
   
	| Code: |  
	| Description: ( WMBBRK6 ) Execution Group '$SYS_mqsi' failed to return an internal configuration 'Start' message response within a 60 second time period.
 
 The internal configuration message response could not be retrieved. The Execution Group therefore failed to move to the state described by the internal configuration message.
 
 Investigate why the Execution Group internal configuration message response was not returned within the allowed time period by reviewing messages issued to the system log. Additionally, you can increase (or decrease) the timeout by using the mqsichangebroker option, ConfigurationDelayTimeout. This will not resolve any underlying problem with a deployed configuration message, but can be used to reduce the response turnaround time or allow for large and complex deployments. Resolve any problems described in the system log. This 'internal configuration' operation will be retried automatically, if you continue to see this message you should investigate the reason for the Execution Groups failure to respond.
 |  |  | 
		
		  | Back to top |  | 
		
		  |  | 
		
		  | jefflowrey | 
			  
				|  Posted: Tue Dec 06, 2005 8:57 am    Post subject: Re: WMB v6 Broker Debug Problems |   |  | 
		
		  | Grand Poobah
 
 
 Joined: 16 Oct 2002Posts: 19981
 
 
 | 
			  
				| 
   
	| TheBigEasy wrote: |  
	| Investigate why the Execution Group internal configuration message response was not returned within the allowed time period by reviewing messages issued to the system log. |  _________________
 I am *not* the model of the modern major general.
 |  | 
		
		  | Back to top |  | 
		
		  |  | 
		
		  | mqmatt | 
			  
				|  Posted: Tue Dec 06, 2005 9:40 am    Post subject: |   |  | 
		
		  |  Grand Master
 
 
 Joined: 04 Aug 2004Posts: 1213
 Location: Hursley, UK
 
 | 
			  
				| I'd start by ensuring a proper default execution group exists. Create one using mqsicreateexecutiongroup and restart the broker. This should get rid of the $SYS_mqsi execution group and we'll be able to see if the problem is related to $SYS_mqsi or not. No reason to suspect why it is, but worth a go.
 |  | 
		
		  | Back to top |  | 
		
		  |  | 
		
		  | TheBigEasy | 
			  
				|  Posted: Wed Dec 07, 2005 5:40 am    Post subject: |   |  | 
		
		  |  Apprentice
 
 
 Joined: 04 Jul 2005Posts: 33
 Location: London
 
 | 
			  
				| Hi Jeff I'd already tried all the obvious investigations - they did not reveal anything of interest. The only other point to note is the broker db threshold alarm being breached by the runaway broker.
 
 Hi Matt
 I did have the default execution group running perfectly fine. The error was the same, I only copied it after I had deleted and tried recreating the broker.
 
 Tried to create a default execution again - however the mqsicreateexecutiongroup connects to the configmgr and then to the broker. In this case the broker is orphaned and consuming memory - the configmgr cannot link to broker (all Qm infrastrucutre is fine).
 
 Resolution:
 Dropped the broker db - deleted the broker and started again. Fairly painless but my confidence is slightly dinted - I do not like not knowing why it went wrong.
 
 Will try to break it again and post the sequence of events.....
   
 Best wishes
 |  | 
		
		  | Back to top |  | 
		
		  |  | 
		
		  | TheBigEasy | 
			  
				|  Posted: Wed Dec 07, 2005 7:39 am    Post subject: |   |  | 
		
		  |  Apprentice
 
 
 Joined: 04 Jul 2005Posts: 33
 Location: London
 
 | 
			  
				| Successfully managed to recreate the problem. 
 1) Debug terminated & removed
 
 2) Broker stopped from the cmd line
 
 3) mqsisetdbparms <brokerName> -n <dbName> -u <userName> -p <password>
 
 4) Broker started from the command line
 
 5) Broker debug session launch attempted
 
 ERROR pop up
 Visual Flow Debugger Launcher
 
 
   
	| Quote: |  
	| Cannot launch this debug session. 
 Reason:
 Flow Engine name is invalid
 or broker has been attached to another debug session
 or broker is not running
 or IBM Agent Controller cannot detect the broker
 |  Comments:
 Flow Engine name is valid - just been used successfully
 broker not attached to another debug session
 broker was restarted after setting the dbparms
 Agent controller was restarted
 
 
 Errors from the Event log - nothing showing in the other logs
 
 
 
   
	| Code: |  
	| Event Type:   Error Event Source:   DB2-0
 Event Category:   None
 Event ID:   4
 Date:      07/12/2005
 Time:      15:19:33
 User:      N/A
 Computer:
 Description:
 
 2005-12-07-15.19.33.667000   Instance:DB2   Node:000
 PID:3832(db2fmp.exe)   TID:3588   Appid:none
 Health Monitor  HealthIndicator::update Probe:500
 
 ADM10500E  Health indicator "Table Space Container Utilization"
 ("tsc.tscont_util") breached the "upper" alarm threshold of "90 %" with value
 "96 %" on "table space container" "DB2.WMBBRK6
 .USERSPACE1.C:\DB2\NODE0000\SQL00001\SQLT0002.0".  Calculation:
 "((os.fs_used/os.fs_total)*100);" = "((37974622208 / 39678746624) * 100)" = "96
 %".  History (Timestamp, Value, Formula): "()"
 
 
 
 Event Type:   Error
 Event Source:   WebSphere Broker v6000
 Event Category:   None
 Event ID:   2080
 Date:      07/12/2005
 Time:      15:20:44
 User:      N/A
 Computer:
 Description:
 ( WMBBRK6 ) Execution Group 'default' failed to return an internal configuration 'Start' message response within a 60 second time period.
 
 The internal configuration message response could not be retrieved. The Execution Group therefore failed to move to the state described by the internal configuration message.
 
 Investigate why the Execution Group internal configuration message response was not returned within the allowed time period by reviewing messages issued to the system log. Additionally, you can increase (or decrease) the timeout by using the mqsichangebroker option, ConfigurationDelayTimeout. This will not resolve any underlying problem with a deployed configuration message, but can be used to reduce the response turnaround time or allow for large and complex deployments. Resolve any problems described in the system log. This 'internal configuration' operation will be retried automatically, if you continue to see this message you should investigate the reason for the Execution Groups failure to respond.
 Data:
 0000: 35 00 30 00 34 00 38 00   5.0.4.8.
 0008: 00 00 57 00 4d 00 42 00   ..W.M.B.
 0010: 42 00 52 00 4b 00 36 00   B.R.K.6.
 0018: 2e 00 61 00 67 00 65 00   ..a.g.e.
 0020: 6e 00 74 00 00 00 46 00   n.t...F.
 0028: 3a 00 5c 00 62 00 75 00   :.\.b.u.
 0030: 69 00 6c 00 64 00 5c 00   i.l.d.\.
 0038: 53 00 30 00 30 00 30 00   S.0.0.0.
 0040: 5f 00 50 00 5c 00 73 00   _.P.\.s.
 0048: 72 00 63 00 5c 00 41 00   r.c.\.A.
 0050: 64 00 6d 00 69 00 6e 00   d.m.i.n.
 0058: 41 00 67 00 65 00 6e 00   A.g.e.n.
 0060: 74 00 5c 00 49 00 6d 00   t.\.I.m.
 0068: 62 00 41 00 64 00 6d 00   b.A.d.m.
 0070: 69 00 6e 00 41 00 67 00   i.n.A.g.
 0078: 65 00 6e 00 74 00 2e 00   e.n.t...
 0080: 63 00 70 00 70 00 00 00   c.p.p...
 0088: 36 00 33 00 31 00 39 00   6.3.1.9.
 0090: 00 00 49 00 6d 00 62 00   ..I.m.b.
 0098: 41 00 64 00 6d 00 69 00   A.d.m.i.
 00a0: 6e 00 41 00 67 00 65 00   n.A.g.e.
 00a8: 6e 00 74 00 3a 00 3a 00   n.t.:.:.
 00b0: 67 00 65 00 74 00 58 00   g.e.t.X.
 00b8: 4d 00 4c 00 52 00 65 00   M.L.R.e.
 00c0: 73 00 70 00 6f 00 6e 00   s.p.o.n.
 00c8: 73 00 65 00 00 00 00 00   s.e.....
 00d0: 00 00
 
 |  |  | 
		
		  | Back to top |  | 
		
		  |  | 
		
		  | mqmatt | 
			  
				|  Posted: Wed Dec 07, 2005 8:12 am    Post subject: |   |  | 
		
		  |  Grand Master
 
 
 Joined: 04 Aug 2004Posts: 1213
 Location: Hursley, UK
 
 | 
			  
				| I'd suggest raising a PMR for this. |  | 
		
		  | Back to top |  | 
		
		  |  | 
		
		  | fjb_saper | 
			  
				|  Posted: Wed Dec 07, 2005 1:10 pm    Post subject: |   |  | 
		
		  |  Grand High Poobah
 
 
 Joined: 18 Nov 2003Posts: 20767
 Location: LI,NY
 
 | 
			  
				| and apply the latest fix pack to the broker ?  |  | 
		
		  | Back to top |  | 
		
		  |  | 
		
		  | EddieA | 
			  
				|  Posted: Wed Dec 07, 2005 3:36 pm    Post subject: |   |  | 
		
		  |  Jedi
 
 
 Joined: 28 Jun 2001Posts: 2453
 Location: Los Angeles
 
 | 
			  
				| 
   
	| Quote: |  
	| the latest fix pack to the broker |  Is there one.
   
 Cheers,
 _________________
 Eddie Atherton
 IBM Certified Solution Developer - WebSphere Message Broker V6.1
 IBM Certified Solution Developer - WebSphere Message Broker V7.0
 |  | 
		
		  | Back to top |  | 
		
		  |  | 
		
		  | TheBigEasy | 
			  
				|  Posted: Wed May 03, 2006 1:45 pm    Post subject: |   |  | 
		
		  |  Apprentice
 
 
 Joined: 04 Jul 2005Posts: 33
 Location: London
 
 | 
			  
				| Sorry for the delay, another topic prompted me to check my earlier postings   
 Solution - I went FP01 and have new hardware - no issues since.
 |  | 
		
		  | Back to top |  | 
		
		  |  | 
		
		  |  |