Hello David, Friends,

Re: FTP and EXITONERROR again!

The FTPFD57 patch includes the following fix:
***************************************************************************
  SR: 5003-395012
FROM: FTPFD02
Fix for JCW not set on commands which fail with a 5xx error message.
***************************************************************************

This was an urgent fix to address "data loss" as the result of the JCW not
being set on several 5xx error messages.  The example of this case of data
loss was "CD" to a non-existent group did not set the JCW for the 5xx error
message returned...   Example:

exitonerror
cd xyz
del *

Prior to the SR 5003-395012 if the "cd xyz" failed returning a 550 error
"550 The last component of the pathname "xyz" does not exist. (CIERR 935)"
did not set the JCW, did not abort the job at "exitonerror" and executed the
next command "del *".  Oop's...   Just deleted files under my root or home
directory...

Now with the fix SR 5003-395012 the JCW is set and exitonerror terminates
the job.

Note: exitonerror is a Boolean and toggles from on to off each time the
command is executed... So, if you have a command you which to ignore the
error state you can toggle exitonerror.

ftp> exitonerror
ExitOnError mode on.  (FTPINFO 33)

{ do commands here you do want error checked }

ftp> exitonerror
ExitOnError mode off.  (FTPINFO 34)

{ do commands here you do not want error checked }

ftp> exitonerror
ExitOnError mode on.  (FTPINFO 33)

{ turn it back on for the remainder of the job }

The fix SR 5003-395012 is included in MPE/iX 6.0 base release and this is
the intended operation of "exitonerror".

Regards,

James Hofmeister
Hewlett Packard
Worldwide Technology Network Expert Center
P.S. My Ideals are my own, not necessarily my employers.


----------------------------------------------------------------David Says--
We've just upgraded to 5.5.6 this past weekend, and included with the update
was...

FTPFD57A  GENERAL FIXES FOR FTP FOR 5.5 PUSH (S Patch).

Now of course, the EXITONERROR FTP command works differently than before
(the comment was made that it now works correctly!).

Whereas before, we could have scripts that performed a MKDIR before
performing a CD into that directory, now the MKDIR will cause the whole
script to abort if that directory already exists.

The only solution I can think of is to not have EXITONERROR set to ON. But
I'm concerned that this will allow 'real' errors to go unnoticed.

Any thoughts, advice, help much appreciated.

Regards,
  David
----------------------------------------------------------------David Says--