Strange tclproc behavior.

[Mike Campbell]

Okay TCL gurus out there.  I’ve got an interesting issue.  

[Cloverleaf 5.7, patch 2 on AIX]

I have an xlate that has a couple of called tclprocs to query an Oracle database.  Things are running fine and dandy.  Then all of a sudden one of the tclprocs throws and error and the messages after that hid the error database.  When I stop and restart the thread all systems are go.  

I’m attaching the tclproc, and the log message is below.

Any help [besides taking the TCL class!] would be appreciated.

Thanks.

Mike C.

[pdl :PDL :ERR /0:  ORM_To_UNMC:02/22/2012 18:35:29] PDL signaled exception: code 1, msg device error (remote side probably shut down)

[xlt :xlat:ERR /0:    ORM_xlate:02/22/2012 18:35:38] [0.0.653722920] Xlate ‘ORM_From_Cerner_To_Cbord.xlt’ failed: Tcl callout error

erroCode: NONE

errorInfo:

hstmt is not open

handle hstmt0 not found

   while executing

“odbc SQLExecDirect $hstmt $queryDB SQL_NTS”

   (procedure “CBORD_Diet_Text_Single_Select_Array” line 86)

   invoked from within

“CBORD_Diet_Text_Single_Select_Array”

[pdl :PDL :ERR /0:  ORM_To_UNMC:02/22/2012 18:35:49] read returned error 0 (Error 0)

Topic

[Jim Kosloskey]

Mike,

I don’t see where you ever check the Sql return and handle SQL_ERROR.

I am guessing you lost connection with the DB.

I also do not see anywhere you are recognizing your connection attempt mayh have been rejected and doing any kind of handling for that eveent (including retry connection).

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

[Mike Campbell]

Looks like the issue was caused by messages hitting the proc too fast and the query wasn’t behaving as expected.  I added a “sleep 2” in the procs and that appears to have fixed the issue.  

  Mike C.

[Author]

Replies