Where can I find newer RTIF proc?

[Tom Patton]

Hi,

I’m running into an issue where I am receiving a SHUT from Invision when waiting on an ACK.  This causes the process to shutdown.  Unfortunately, the SHUT msg is still kept in state 1 and when the process is brought back up, it processes this msg again and the process goes down again.

I had heard (in the old archive) there was a newer proc.

I wonder if I can get a copy or if anyone has some ideas on a correction. I have some ideas like altering the msg and killing it when the process runs again.

Thanks,

Topic

[James Cobane]

Tom,

I’m not sure if there are any updated procs from Quovadx, but perhaps someone has modified the RTIF procs on their own.  The issue with this is that the engine is invoking the hcicmd command to stop the process within the same tcl stream, thus the message gets ‘bound in tcl’ in state 1 and hence gets reprocessed each time you re-start the process.  Ultimately, you have to delete the 4SHUT message from the recovery database, then start the process.  Good Luck.

Jim Cobane

Henry Ford Health

[Charlie Bursell]

Good point Jim.  I can’t tell you how many TIF/RTIF procs (and others) I have seen out there still stating they were written by me when the code does not even resemble what I originally wrote.

All I can say is beware of what is passed to you by another user.  Just because it works for them, does not mean it will work for you.  Seems a lot of people go broke trying to save money  😆

With that said, I sent him my latest version.

[Tom Patton]

I should have posted a follow-up.  Jim, that was the issue from what I can tell.  Charlie sent me another copy of the RTIF that had a:

rtif::shutdown

return “{KILL $mh}”

instead of:

rtif::shutdown

return “{}”

I wasn’t certain if the cmd process would still be able to kill the msg after the shutdown, and knew the proc had been modified, so I went looking.

And I agree on the proc status.  I inherited this one and it has a few other flaws.

One weakness is that it doesn’t handle is the Invision “last record / not last record” indicator in the header.  We are working on adding the functionality as well as variable padding for the header to allow for varying header size requirements.  This will all be done via arguments in a proc name other than “RTIF” – to not add to the confusion.

Thank you for the note.

[Author]

Replies