filePro data entry failing on restored 5.6.10D4 Linux server

Mike Schwartz mschw at athenet.net
Thu Jan 28 08:01:38 PST 2016


     Last weekend one of my customers had a hard drive fail on their Red Hat
Enterprise Linux 6.7 server.  They are running filePro 5.6.10D4 (Linux).

 

     Fortunately, I had them purchase an exact duplicate HP DL120 G5 server
a few years ago.  I had done a test restore of their entire system onto this
backup server using Microlite's Backup Edge, so I thought we were fully
prepared for any catastrophic failures.

 

     I popped the backup server into their server rack and did a full Backup
Edge restore off of their FTP server.  Everything appeared to restore
flawlessly.  I had some difficulty re-licensing their filePro because
filePro's license manager was only detecting one of the 3 MAC addresses on
that machine, but eventually, thanks to a lot of troubleshooting by Ernie,
we got the filePro re-licensed.

 

     I thought the job was done and I was just about to pat myself on the
back, when their data entry people started coming to me and saying they were
getting "kicked back" to the filePro menus in the middle of entering data.
The "kick backs" seemed to be random; not related to using a particular key
or entering data into a specific file.

 

   Some users are using a Putty terminal emulator and some are using
Ice-TCP, both using SSH, so it's not related to just one terminal emulation
program.

 

   The users are not losing their connection to the server.  Some menus are
4 or 5 menus deep and they said the system is "jumping them back to a
different menu each time", but I suspect that, since they are all high-speed
touch-typists, they are continuing to type a few characters after they get
kicked out.   

 

    Their network appears to be rock solid; professionally installed cables
and 1-year old HP switches.  I can do full master Backup Edge backups and
verified restores from this server to their FTP server.  Loop tests allowed
me to FTP a large file 1,000 times between my laptop and the server without
any dropped data.    

 

----- ----- -----IMPORTANT:

 

    I asked one user to do some of her timecard data entry DIRECTLY ON THE
SERVER CONSOLE, and she got "kicked back" after entering just a couple of
timecards, just like she did on her workstation using Ice-TCP, so I think I
have ruled out any network issues.  This has to be something to do  with
filePro, itself!  I can browse the Internet and do everything else I could
normally do on a Linux server on this computer.  (At least I haven't found
anything else that is NOT working...)

----- ----- --------------------

 

    1) I ran my filePro reindex script and it appeared to rebuild all the
indexes.  There were a couple of errors that flew past that I didn't catch,
but I am going to rebuild the indexes on some of the files manually after I
finish this email.  

 

    2)  I ran my "rebuild freechains" script on all the filePro files.  It
appeared to run correctly.  

 

    3)  I removed all the lockfiles on all the filePro files.  

 

    4)  Fortunately, the hard drive apparently failed on Sunday, so no
filePro files would have been open at the time of the failure and the Friday
evening backup captured all the data that had been entered on Friday, so no
data was lost.

 

     I'd appreciate any ideas, because this customer will have to shut down
their plant (100 people) if I can't get them up and running VERY SOON!

 

    I'm going to get cleaned up and drive out there with new hard drives for
their OLD server this morning.  I will try to get 

 

Thanks!

 

Mike Schwartz

 

  

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.celestial.com/pipermail/filepro-list/attachments/20160128/7222abd0/attachment.html>


More information about the Filepro-list mailing list