RE: ICF file error "transmission end" during read

I am still researching ICF programming as it is new to me.
The major code was 81 and the minor code was 97 from the dump.  This is not
unexpected though as the file was not on the remote system and I would
receive these same errors the old way.  The difference though is that the
old way (RPG) would turn on the indicators on the "read" rather than
blowing up (RPGLE).  I have indicators for LO and EQ but not HI.  Just like
in the old RPG.  I ran the new way and the old way many times over a couple
days.  In every case, the new way would error out but the old way would
complete fine.  The dump showed it blowing on the read where it was trying
to read the data from the remote system.  It seems like RPGLE blows on this
major code but RPG does not.  Someone mentioned an ADDICFDEVE.  I think the
overrides are fine for the device as seen at the bottom.  I do not need the
ADDICFDEVE as that occurs in the OVRICFDEVE.  I know it was on the read of
the data because I could see 2 puts and 2 gets in the dump that were from
the dialing and logging on.  The input and output buffer contained the
logon userid and password.  Any ideas?
I was able to get a contact so I can probably be able to get a test setup.
I saw mentioned in this list, CRTDEVINTR for testing on the same system.
I'm not sure how to setup how to get to what file so I will probably see if
I can get a test with the bank.
Thanks for the responses so far!  I will try to keep you posted.  Sorry
about the long delay.

Craig Strong

*** Mike wrote:

Craig,

Does this occur everytime with your ILE version, or just the one time?

Also, what information was available at the RPG layer (i.e. major/minor
codes)?  Is it possible that you received a "forward abort sequence" (EOT
before ETX) from the remote that wasn't handled by your program?  While
the BSC protocol is fairly basic,  there is still a conversation sequence
that must occur by definition (i.e. SOT, ETX, TTD, EOT, etc....).  I have
seen where poor line quality can cause such behavior.

Regards,

Michael Rooney
Citigroup International







-----Original Message-----
From: rpg400-l-bounces@xxxxxxxxxxxx
[mailto:rpg400-l-bounces@xxxxxxxxxxxx]On Behalf Of craigs@xxxxxxxxx
Sent: Wednesday, June 16, 2004 2:39 PM
To: rpg400-l@xxxxxxxxxxxx
Subject: ICF file error "transmission end" during read






Does anyone know what might cause a "Transmission end received from device
<device> for file <ICF file> in <library>"?
More details: " A BSC transmission end sequence  was received from
device...".
Technical description: Error resource name <line>, hardware error code
92080000.... Hardware error code 92080000, 93880000.

I was trying to receive a file from a bank and I received this error during
the read of the ICF file.  Somehow it corrupted and deleted the file from
the bank.  Luckily this was basically a clone of an old version that I
copied and converted to RPGLE.  So, the bank just set the file back out
there and we ran the old version and it received fine.
The new version is for switching between FTP and modem easily.  Just call a
different string of programs with similar names.  From what I can tell from
the statement number, it dialed fine, logged on fine, but then errored
during receival of the file (read RCVDATA with Lo ind=98 and Eq ind=99).
In the ICF file, RCVDATA record is defined with RCVFLD 500 and record
keywords RCVTRNRND(81) TEXT('RECEIVE RECORD').
I compared old and new ICF file descriptions (with DSPFD), old and new
programs, etc and I do not see any differences.
I still do the OVRICFF FILE(<file>) ACQPGMDEV(<device>) LVLCHK(*NO).
Then OVRICFDEVE PGMDEV(<device>) RMTLOCNAME(<rmtloc>) CMNTYPE(*BSCEL)
FMTSLT(*PGM) BLOCK(*DEVD) BLKLEN(*DEVD) TRNSPY(*DEVD) DTACPR(*DEVD)
RMTBSCEL(*NO) INLCNN(*DIAL).
Then two VRYCFG for *LIN and *CTL to STATUS(*ON).

I cannot see anything that has changed or that may be hardcoded.  The FTP
test works fine but I still want to have modem as a backup.  I am waiting
to be able to test on the bank's development system as I am scared to
affect production data.  I'm not sure if we could setup a test on our
development system or not.  Any ideas about this problem?

Thanks!
Craig Strong