data TPS: external proc ‘cljTPS’ not available

[Doug Miller]

Cloverleaf 6.1.1

Trying to implement a known working Java class, but get the following error upon validation in Net Config:  data TPS: external proc ‘cljTPS’ not available.  Getting the errors below after trying to start the thread.  Using same procedure and configuration as with other 6.1.1 systems where this is working without issue.  Permissions all seem to be set correctly, but this system can’t seem to recognize the proc.  Any suggestions?

[sms :sms :ERR /0: CPACS_ENM_In:09/21/2018 15:10:46] Tcl error:

[sms :sms :ERR /0: CPACS_ENM_In:–/–/—- –:–:–]          msgId    = none

[sms :sms :ERR /0: CPACS_ENM_In:–/–/—- –:–:–]          proc      = ‘cljTPS’

[sms :sms :ERR /0: CPACS_ENM_In:–/–/—- –:–:–]          args      = ‘{CLASS CCGFilterDuplicateA08_v9x} ‘

[sms :sms :ERR /0: CPACS_ENM_In:–/–/—- –:–:–]          result    = ‘invalid command name “cljTPS”‘

[sms :sms :ERR /0: CPACS_ENM_In:–/–/—- –:–:–]          errorInfo: ‘

[sms :sms :ERR /0: CPACS_ENM_In:–/–/—- –:–:–] invalid command name “cljTPS”

[sms :sms :ERR /0: CPACS_ENM_In:–/–/—- –:–:–]     while executing

[sms :sms :ERR /0: CPACS_ENM_In:–/–/—- –:–:–] “cljTPS {CONTEXT sms_ib_data} {ARGS {{CLASS CCGFilterDuplicateA08_v9x} }} {MODE start} {VERSION 3.0}”‘

Topic

[Rob Abbott]

Your root tclIndex is probably corrupted.

Run the “mktclindex” command line in $HCIROOT/tclprocs to fix it.

Rob Abbott
Cloverleaf Emeritus

[Doug Miller]

Great thought, but that didn’t correct the issue.  The Java class file is stored in the sitejava_uccs folder.  For the other servers where this is working, we just copy the .class file to the site java_uccs folder and it’s ready to go.  In this case, that doesn’t seem to be the case.  I have checked permissions and all the correct users appear to have full permissions to the .class file and the java_uccs folder.

[David Barr]

You probably need to restart your engine process, and you might need to restart the host server if you’re using the gui to restart processes.

[Doug Miller]

I have tried that in the past also, but just tried each of those steps again along with rebooting the server – same error.

[Tim Pancost]

Frankly, it would be my guess that what Rob suggested to do to fix the issue(run mktclindex in $HCIROOT/tclprocs) is what caused the error in the first place.  That is, someone ran mktclindex in your $HCIROOT/tclprocs directory.  You cannot run mktclindex in $HCIROOT/tclprocs for that very reason, it removes the references to the java objects(cljTPS, cljXLTStrings, cljXLTObjects, cljTrxID, cljRMThreadAssoc, etc… basically everything in $HCIROOT/tcl/lib/tfc/cloverleafJava.tcl except some j-stuff) from the tclIndex file, as mktclindex only looks in the current directory when it’s creating the tclIndex file.

I’m guessing that if you looked in your $HCIROOT/tclprocs/tclIndex file, you’ll find the following entry missing:

set auto_index(cljTPS)

]

As for how to correct it, we’ve historically just taken the $HCIROOT/tclprocs/tclIndex file from one of our other machines and replace the one that was corrupted.  If you don’t have another machine, hopefully you have a backup strategy from which to pull.

HTH,

TIM

Tim Pancost
Trinity Health

[Doug Miller]

That was exactly the piece of information I needed, thanks!  I simply edited the tclIndex and .upocindex to add the missing line for cljTPS.  That seems to have corrected the problem.  Messages are passing and the script is working now.

😀

[Rob Abbott]

I stand corrected!  Thank you Tim!

Rob Abbott
Cloverleaf Emeritus

[Author]

Replies