| Author |
Message
|
| Abhinay185127 |
 Posted: Fri Jul 26, 2013 4:07 am Post subject: Reg : Subflow |
 |
|
Acolyte
Joined: 20 May 2013
Posts: 58
|
Hi
I am using a subflow in my mainflow. This subflow i am calling four times to log messages.
for the first time it works fine after that when second time i am invoking it is not able to invoke that. I am using subflow as .subflow.
Any idea where i am going wrong |
|
| Back to top |
|
 |
| lancelotlinc |
| Posted: Fri Jul 26, 2013 4:11 am Post subject: |
|
|
Jedi Knight
Joined: 22 Mar 2010
Posts: 4941
Location: Bloomington, IL USA
|
Your missing some essential details in your post:
1. Version of toolkit.
2. Version of runtime.
3. Effective level of runtime.
4. Diagram of your main flow.
5. Do you have Trace nodes placed in proper places?
_________________
http://leanpub.com/IIB_Tips_and_Tricks
Save $20: Coupon Code: MQSERIES_READER |
|
| Back to top |
|
|
| mqjeff |
| Posted: Fri Jul 26, 2013 4:59 am Post subject: |
|
|
Grand Master
Joined: 25 Jun 2008
Posts: 17447
|
| lancelotlinc wrote: |
Your missing some essential details in your post:
1. Version of toolkit.
2. Version of runtime.
3. Effective level of runtime.
4. Diagram of your main flow.
5. Do you have Trace nodes placed in proper places? |
Most of this is going to end up being completely irrelevant to the problem trying to be solved.
It's still a useful exercise to gather, in that if the toolkit and runtime levels are not at 8.0.0.2, then using a .subflow is fraught with peril.
Abhinay185127 - you have not given a good description of what is going on. Please explain exactly what it means that "it is not able to invoke that". How do you know this? What data shows you this? was it a user trace data? Was it you merely staring at your code and guessing wildly?
What process of troubleshooting have you followed to determine that this is the problem? |
|
| Back to top |
|
|
| lancelotlinc |
| Posted: Fri Jul 26, 2013 5:22 am Post subject: |
|
|
Jedi Knight
Joined: 22 Mar 2010
Posts: 4941
Location: Bloomington, IL USA
|
| mqjeff wrote: |
| lancelotlinc wrote: |
Your missing some essential details in your post:
1. Version of toolkit.
2. Version of runtime.
3. Effective level of runtime.
4. Diagram of your main flow.
5. Do you have Trace nodes placed in proper places? |
Most of this is going to end up being completely irrelevant to the problem trying to be solved.
It's still a useful exercise to gather, in that if the toolkit and runtime levels are not at 8.0.0.2, then using a .subflow is fraught with peril.
Abhinay185127 - you have not given a good description of what is going on. Please explain exactly what it means that "it is not able to invoke that". How do you know this? What data shows you this? was it a user trace data? Was it you merely staring at your code and guessing wildly?
What process of troubleshooting have you followed to determine that this is the problem? |
I disagree. If on 8.0.0.1, there is a known issue with subflow invocation. If there are no Trace nodes, how does the OP know that the subflow is not getting invoked? If there are no Trace nodes, how does the OP know that an exception path is not being triggered?
It is entirely relevant.
_________________
http://leanpub.com/IIB_Tips_and_Tricks
Save $20: Coupon Code: MQSERIES_READER |
|
| Back to top |
|
|
| smdavies99 |
| Posted: Fri Jul 26, 2013 5:34 am Post subject: |
|
|
Jedi Council
Joined: 10 Feb 2003
Posts: 6076
Location: Somewhere over the Rainbow this side of Never-never land.
|
| lancelotlinc wrote: |
If there are no Trace nodes, how does the OP know that an exception path is not being triggered?
|
You most certainly DO NOT NEED trace nodes to tell if an exceptionPath has been triggered. any system with half decent error handling will be using other methods entirely. Do you really want to have potentially huge disk gobbling trace logs in production etc etc?
Yes they (Trace nodes) are useful but you make it sound to beginners that the only way to find this stuff out is to use a trace node.
Oh dear, perhaps I need re-training as I am clearly in need of it 
_________________
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 |
|
|
| lancelotlinc |
| Posted: Fri Jul 26, 2013 5:36 am Post subject: |
|
|
Jedi Knight
Joined: 22 Mar 2010
Posts: 4941
Location: Bloomington, IL USA
|
| smdavies99 wrote: |
| lancelotlinc wrote: |
If there are no Trace nodes, how does the OP know that an exception path is not being triggered?
|
You most certainly DO NOT NEED trace nodes to tell if an exceptionPath has been triggered. ant system with half decent error handling will be using other methods entirely. Do you really want to have potentially huge disk gobbling trace logs in production etc etc?
Yes they (Trace nodes) are useful but you make it sound to beginners that the only way to find this stuff out is to use a trace node.
Oh dear, perhaps I need re-training as I am clearly in need of it |
Obviously the OP's configuration does not meet your qualification: "ant system with half decent error handling".
Also, production systems usually turn off Trace nodes with the mqsichangetrace command.
Trace nodes are invaluable. There is no reason not to use them.
_________________
http://leanpub.com/IIB_Tips_and_Tricks
Save $20: Coupon Code: MQSERIES_READER |
|
| Back to top |
|
|
| Vitor |
| Posted: Fri Jul 26, 2013 5:38 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| lancelotlinc wrote: |
| If there are no Trace nodes, how does the OP know that the subflow is not getting invoked? If there are no Trace nodes, how does the OP know that an exception path is not being triggered? |
Because the OP will run a user trace which will show the logic path the flow is following. Running a user trace does not require the insertion of trace node, or indeed any flow editing at all.
Didn't they cover user trace in your training course?
| lancelotlinc wrote: |
| It is entirely relevant. |
Only to your various personal crusades. Which are becoming increasingly irrelevant to the posts you're responding to.
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| lancelotlinc |
| Posted: Fri Jul 26, 2013 5:48 am Post subject: |
|
|
Jedi Knight
Joined: 22 Mar 2010
Posts: 4941
Location: Bloomington, IL USA
|
| Vitor wrote: |
| lancelotlinc wrote: |
| If there are no Trace nodes, how does the OP know that the subflow is not getting invoked? If there are no Trace nodes, how does the OP know that an exception path is not being triggered? |
Because the OP will run a user trace which will show the logic path the flow is following. Running a user trace does not require the insertion of trace node, or indeed any flow editing at all.
Didn't they cover user trace in your training course?
| lancelotlinc wrote: |
| It is entirely relevant. |
Only to your various personal crusades. Which are becoming increasingly irrelevant to the posts you're responding to. |
If you don't want to use Trace nodes, don't use them. User trace is sometimes unavailable in certain environments due to local policies.
Are we throwing mud this morning Vitor ?
I'm not a crusader. If someone asks "Any idea where i am going wrong" my response is to add visibility into the inner workings of the OPs flow. One way to do that is Trace nodes. In fact, the InfoCentre documents this technique:
http://publib.boulder.ibm.com/infocenter/wmbhelp/v8r0m0/index.jsp?topic=%2Fcom.ibm.etools.mft.doc%2Fac16810_.htm
| Quote: |
Viewing the logical message tree in trace output
To view the structure of the logical message tree at any point in the message flow, include a Trace node and write some or all the message (including headers and all four message trees) to the trace output destination.
You might find trace output useful to check or record the content of a message before and after a node has changed it, or on its receipt by the input node. For example, if you include a Compute node that builds a destination list in the local environment tree, you might want a record of the structure that it has created as part of an audit trail, or you might want to check that the Compute node is working as you expect it to. |
And the InfoCentre documents how to turn off Trace nodes in Production environments:
http://publib.boulder.ibm.com/infocenter/wmbhelp/v8r0m0/index.jsp?topic=%2Fcom.ibm.etools.mft.doc%2Fau60100_.htm
| Quote: |
| Switching Trace nodes on and off |
If you believe this information is incorrect or needs revising, you can submit feedback on the page indicating your disagreement with documented practices.
_________________
http://leanpub.com/IIB_Tips_and_Tricks
Save $20: Coupon Code: MQSERIES_READER |
|
| Back to top |
|
|
| Vitor |
| Posted: Fri Jul 26, 2013 6:07 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| lancelotlinc wrote: |
| If you don't want to use Trace nodes, don't use them. User trace is sometimes unavailable in certain environments due to local policies. |
Or available only in response to a production ticket. What other "local policies" would prevent use of user trace but allow the use of Trace nodes (when enabled using the techniques you describe).
| lancelotlinc wrote: |
| Are we throwing mud this morning Vitor ? |
Are you worried about getting mud on your shiny armour or on your white charger?
| lancelotlinc wrote: |
| I'm not a crusader. If someone asks "Any idea where i am going wrong" my response is to add visibility into the inner workings of the OPs flow. One way to do that is Trace nodes. |
Which you do not mention. You espouse Trace nodes as the only solution. In this context they do not help the OP's ability to determine flow. For other problems, such as parser or data issues, Trace nodes are invaluable.
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| smdavies99 |
| Posted: Fri Jul 26, 2013 6:25 am Post subject: |
|
|
Jedi Council
Joined: 10 Feb 2003
Posts: 6076
Location: Somewhere over the Rainbow this side of Never-never land.
|
| Vitor wrote: |
Which you do not mention. You espouse Trace nodes as the only solution. In this context they do not help the OP's ability to determine flow. For other problems, such as parser or data issues, Trace nodes are invaluable. |
 
That's what I was implying. Mr <man on white charger> gives makes lots of posts telling people to use a certain set of setting in trace nodes. In many instances these suggestions are next to useless while in others they are spot on the mark.
Then there are the issues around the effective runtime operating level when the problem is in the tooklit.
Stock response No 3 is get some training, which I clearly need  because I dare to do some things differently to others.
It is Friday and nearly Beer O'clock. Hurrah. Hair of the hog tonight.
_________________
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 |
|
|
| lancelotlinc |
| Posted: Fri Jul 26, 2013 7:05 am Post subject: |
|
|
Jedi Knight
Joined: 22 Mar 2010
Posts: 4941
Location: Bloomington, IL USA
|
| Vitor wrote: |
| You espouse Trace nodes as the only solution. |
I never said they were the only option.
| Vitor wrote: |
| In this context they do not help the OP's ability to determine flow. |
This is entirely false. Trace nodes give supreme visibility into the flow order, which path gets invoked when, and perhaps when an exception route is taken unexpectedly.
| Vitor wrote: |
| For other problems, such as parser or data issues, Trace nodes are invaluable. |
The disadvantage of user trace is (in addition to being restricted by local policy) one must sift through a large output. With Trace nodes, you simply sort on file changed date in your explorer to see which path was taken when.
Again, no one is forcing you to use anything. You use what is convenient for you. If you don't like Trace nodes, don't use them. But to say that Trace nodes are irrelevant is entirely false.
And I don't see any information from either of you two helping the poster. mqjeff started on a good tact. We'll see how the troubleshooting progresses.
In the end, any tool that can give visibility into the logic the message flow is following is a good thing.
_________________
http://leanpub.com/IIB_Tips_and_Tricks
Save $20: Coupon Code: MQSERIES_READER |
|
| Back to top |
|
|
| lancelotlinc |
| Posted: Fri Jul 26, 2013 7:09 am Post subject: |
|
|
Jedi Knight
Joined: 22 Mar 2010
Posts: 4941
Location: Bloomington, IL USA
|
| smdavies99 wrote: |
| It is Friday and nearly Beer O'clock. Hurrah. Hair of the hog tonight. |
I suspect someone has been imbibing already.
_________________
http://leanpub.com/IIB_Tips_and_Tricks
Save $20: Coupon Code: MQSERIES_READER |
|
| Back to top |
|
|
| Vitor |
| Posted: Fri Jul 26, 2013 7:59 am Post subject: |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| lancelotlinc wrote: |
| I never said they were the only option. |
When do you mention other options? How many time (by comparison) do you ask a poster where the Trace nodes are in a flow, implying that they should be present and should be added if absent?
| lancelotlinc wrote: |
| Vitor wrote: |
| In this context they do not help the OP's ability to determine flow. |
This is entirely false. Trace nodes give supreme visibility into the flow order, which path gets invoked when, and perhaps when an exception route is taken unexpectedly. |
So is your statement. Supreme visability into the logic path of a flow is given by the user trace, which will show the path being taken irrespective of the position of any trace nodes and also gives valuable information on non-exception brancing performed by the flow.
| lancelotlinc wrote: |
| The disadvantage of user trace is (in addition to being restricted by local policy) one must sift through a large output. With Trace nodes, you simply sort on file changed date in your explorer to see which path was taken when. |
I again ask what local policies would prevent the use of user trace but allow the use of Trace node. I repeat my comments above on the ability of user trace to describe non-exception decisions, and point out you're not using a Trace node to display message tree per se, but are using it like a COBOL DISPLAY statement to output a file name at a given point as an eye catcher.
| lancelotlinc wrote: |
| Again, no one is forcing you to use anything. You use what is convenient for you. If you don't like Trace nodes, don't use them. But to say that Trace nodes are irrelevant is entirely false. |
I believe I gave a use case where I described them as "invaluable".
| lancelotlinc wrote: |
| And I don't see any information from either of you two helping the poster. |
We're both suggesting taking a user trace, which I took to be one of the other options which you alude to when you say:
| lancelotlinc wrote: |
| I never said they were the only option |
If your assertion is that we're not suggesting anything which helps the OP, what other options are you in fact refering to?
| lancelotlinc wrote: |
| I suspect someone has been imbibing already |
For the record, I'm stone cold sober. So:
- what "local policies" allow Trace nodes but not user trace?
- what are the "other options" as user trace is in your view not advice which is helping the OP?
- why do you need to "sift" through user trace when any text editor can find the needed section instantly?
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| smdavies99 |
| Posted: Fri Jul 26, 2013 9:06 am Post subject: |
|
|
Jedi Council
Joined: 10 Feb 2003
Posts: 6076
Location: Somewhere over the Rainbow this side of Never-never land.
|
| Vitor wrote: |
| lancelotlinc wrote: |
| I suspect someone has been imbibing already |
For the record, I'm stone cold sober. So:
|
So am I as I'm driving for the next 90 mins. Then I'll have a Pint.
What Vitor and I were alluding to was the fact that there seems to be three canned replied in many of the posts you answer.
1) Get some training
2) Runtime version
3) Trace Nodes.
There is as you well know, a lot more to broker fault finding and problem solving and any one or even all three of the above. Sure, those answers have a part to play when they are relevant but surely not on every post?
Sometimes it appears that you post just for the hell of it. This is probably not the case and that you have only the best intentions but from where I'm sitting (not in a pub/bar) is looks a lot different.
_________________
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 |
|
|
| lancelotlinc |
| Posted: Fri Jul 26, 2013 9:12 am Post subject: |
|
|
Jedi Knight
Joined: 22 Mar 2010
Posts: 4941
Location: Bloomington, IL USA
|
| Vitor wrote: |
- what "local policies" allow Trace nodes but not user trace?
- what are the "other options" as user trace is in your view not advice which is helping the OP?
- why do you need to "sift" through user trace when any text editor can find the needed section instantly? |
In my current environment (the major equipment manufacturer of the Internet), DEV is wide open, I can do anything I want. QA is closed to me, but open to SysAdmin. Prod would be difficult to get anything but the (log4j) logs from. If the OP wants to take a user trace, more power to him.
In banking and insurance (my last two assignments), developers are explicitly prohibited from accessing any PROD artifact.
_________________
http://leanpub.com/IIB_Tips_and_Tricks
Save $20: Coupon Code: MQSERIES_READER |
|
| Back to top |
|
|
|
|
