message ID, no error

[Paul Sullivan]

Is there a way to see a message based on the message ID that didn’t produce an error. I have a warning in the process log referring to a segment out of order. All I have to go on is ORU and the message ID.

Topic

[Jim Kosloskey]

Paul,

Try looking in the SMAT file.

email: jim.kosloskey@jim-kosloskey.com 29+ years Cloverleaf, 59 years IT - old fart.

[Paul Sullivan]

Can I search on the 0.0.xxxxxx message ID in SMAT? There are many routes in this one process, makes it rough to pin it down.

[David Barr]

You can use grep on the .idx files with the message ID (without the 0.0. prefix) to identify which saved message file contains the message, then use SMAT to search that file for the message ID.

[Paul Sullivan]

Did a grep on *.idx. Came up with nothing. This is from today, so I should not need to worry about an archived message.

[Jim Kosloskey]

Paul,

Make sure you are searching the inbound SMAT file from the Source thread.

If you have intermediate threads between the inbound thread and where the Xlate is done your source thread will be one of those intermediate threads.

email: jim.kosloskey@jim-kosloskey.com 29+ years Cloverleaf, 59 years IT - old fart.

[Paul Sullivan]

Looking more into this. I have a notion that our SMAT files are not saving any message ID’s. Tried using the message ID range search in SMAT on some messages from today and got nothing. Is there somewhere SMAT determines what is saved.

[Jim Kosloskey]

Message ID is always saved in the .idx file of the SMAT set so you should be able to find it.

You will note it is displayed in the SMAT display in the ‘Current Message’ Section when you are viewing a message.

email: jim.kosloskey@jim-kosloskey.com 29+ years Cloverleaf, 59 years IT - old fart.

[Paul Sullivan]

Sorry to keep at this but, I looked through random SMAT files and saw the message ID window. I did a grep on those ID’s and had a hit. Below though is a snipit from the process .err log. Neither of these ID’s respond to grep.

[msg :RecP:WARN/0:    rad_xlate:04/08/2014 07:26:13] [0.0.53146388] The # 6

ent encountered ‘DG1’ is out of order for message type ‘ORU_R01’.  Segment

ed.

[xpm :xlt :WARN/0:    rad_xlate:04/08/2014 07:26:13] Warning during PathCop

e # 6 segment encountered ‘DG1’ is out of order for message type ‘ORU_R01’.

ment ignored.

[msg :RecP:WARN/0:    rad_xlate:04/08/2014 07:40:07] [0.0.53148712] The # 7

ent encountered ‘DG1’ is out of order for message type ‘ORU_R01’.  Segment

ed.

[/b]

[Bob Richardson]

Grretings,

I usually only use message ID when dumping explicit messages from either the Error or Recovery databases as the engine keeps changing the message ID as it moves the message from queue to queue internally.

To find a message that is causing errors or other problems, I usually located the date and time in the process log where the warning or error is displayed and then search SMAT for messages in that date and time range.  Then run the selected subset through a route test to see if the same error or warning displays crop up.

In other words, trying to find an exact message ID that is echoed in a process log may never yield a successful find due to the internal movement of the message.

You can demonstrate that for yourself by turning up the engine noise (EO filter) and track how a message gets reassiged a new message ID.

Granted:  more work but so far as proven effective.

Hope this helps you out.

[Author]

Replies