Filepro-list Digest, Vol 50, Issue 30

[Tyler]

>
> Message: 4
> Date: Thu, 13 Mar 2008 17:19:06 -0400
> From: Nancy Palmquist <nlp at vss3.com>
> Subject: Re: fpCGI failure
> To: Tyler Style <tyler at healthyhabitsweb.com>
> Cc: filepro-list at lists.celestial.com
> Message-ID: <47D99A4A.3050906 at vss3.com>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
> Tyler Style wrote:
> > It would be very odd for one report bombing out to block every
> > subsequent report from running; if that was the case, it should bring
> > our whole system to a halt right away as no one would be able to run
> > reports at all.  I think this isolates it pretty much to the actual
> > fpcgi binary.
>
> Not odd at all.  I see just this behavior when something gets stuck and I
> hit
> the license limit.  Then everything gets stuck.
>
>
> The hung process is waiting for something.


Ah.  Licensing is definitely not the issue here, I am certain of that.  And
the process is not hung; there are no entries in the process table for it.
Besides, what would processing like

::html :cr @pw{".htm"
::html :tx "running"
::html :cr-
::exit

be waiting on?  No, we're pretty sure it's the binary.

Feature of fpcgi - look in those docs.  Ask Ray to send you current ones if
> you
> don't have time.  This was new to 2.0.


Yah, fptech hasn't put it in the online manual,  and our reseller never gave
us the dox.  We'll have to see about getting some.

Is this 5.6 and if it is, did someone tell you how to check to see if the
> licensing was exceeded?  I have 5.6 but have not used it enough to
> remember how
> to handle the tools.


5.0.14, and no licensing issues.


> I had issues like this with filePro.  It would exit with a Seg Violation.
>  The
> Return code 0 tells you nothing.  These errors were buggers to find.  I
> added
> log entries to my processing using LOGTEXT so I could narrow down to where
> the
> crash was happening.


We've definitely had our share of those!  :P  They are no fun.

I can honestly say that it could be hundreds of things but here is my check
> this
> list:
>
> 1) make sure you are not using a Declared Variable that matches a Field
> Heading.
> 2) make sure you lock and write all records that you change.
> 3) make sure all variables are defined correctly.
> 4) check your labels and make sure they do not match declared variables,
> array
> names, commands, or aliases.
>
> I fussed with stuff for months until I finally found it.  I think the
> suggestions about writing stdout and errout to files helped.  Also running
> it
> outside of fpcgi, in a test mode, pointed out a few issues.
>
> My tables use many calls, Open(), import/export, all those things, and
> when
> tables are layered like that it is easy to tromp on something.
>
> Once I got it fixed, it has run flawlessly for months.  I know it can
> work, the
> error was mine.  It would have been nice if filePro could have helped with
> better syntax checking or something but the responsibility is all mine.


As the processing table above shows, it is definitely not a processing table
issue.   There's nothing there for it to die on.  Also, most of our
processing doesn't use any HTML in filepro at all.  We pump out JSON
formatted datagrams to AJAX mostly.

I works exactly as you indicated, when it is broken.  You set variables to
> tell
> filePro to continue after x seconds, so it just fails and fpcgi returns
> nothing
> since filePro made nothing.  fpCGI is very happy, you are not back at the
> web
> page since nothing is there.


> BROKEN submission:
> ----------------
> 2008-03-06 14:22:26 INFO   PID(00008186) SUCCESSFUL return from command
designated in Field_cmd
> 2008-03-06 14:22:26 INFO   PID(00008186) Command string is
/appl/fp/rreport ajax -fp cWebInvoice -sr 34 -n -u -y automaticCGI -rw
/usr/local/apache/htdocs
>
> /rweb2a2
> > 2008-03-06 14:22:26 INFO   PID(00008186) Return code from command is 0.
> Errno is 0
> > 2008-03-06 14:22:26 INFO   PID(00008186) Command as returned by
> cgival(Field_cmd1) is L
> > <snip>
> > 2008-03-06 14:22:26 DEBUG  PID(00008186) fpcgi.c 001553 Checking to see
> if [/usr/local/apache/htdocs/rweb2a2.htm] exists
> > 2008-03-06 14:22:26 DEBUG  PID(00008186) fpcgi.c 001573 No. It does not
> exist. Errno is 2


That's the problem, it IS broken there - there SHOULD be output.  If it is
returning without an error, then like the 1200 trial runs before it, there
should be output.  That exact command was run flawlessly hundreds of times
before, with exactly the same data.  Suddenly, it is not pumping out files
anymore, and every subsequent process that runs via fpcgi after that also
fails to output anything at all, regardless of the command being run.

Tyler
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.celestial.com/pipermail/filepro-list/attachments/20080313/5677cbc3/attachment.html