next message in archive
no next message in thread
previous message in archive
previous message in thread
Index of Subjects
Index of Subjects
Hello All!
Further testing on zmodem reveals I know not what :)
At 07:31 Mar 7 approx., on aa146, I began a zmodem download of the
andrew word processor from the PDA, approx 1.5 mb.
93% complete, (1.4 mb, 07:46) the line HUNG UP. The comm program
thought things were still in progress, i.e. the download dialog was
still open. There had been a bit of activity on the send data
indicator on the modem (USR Sportster 14400) just before hangup.
pulling .shell-errors by ftp from aa146:
bash$ cat .shell-errors
/local/bin/sz2: 10387 Hangup
Terminated
So sz2 either detected a hangup, or forced one.
Questions:
(1) does zmodem assume packets are OK unless a NAK is returned?
(looks like it)
(2) does sz2 not properly handle some NAK's?
(3) does sz2 generate a sig when it tries to handle some exception
that we are not trapping? (check source for above error code)
(4) was the hangup generated on the server side or the client side?
(again, source for error code probably will tell)
(5) what did the idle-log say about the termination of that session?
(David Potter, can you check?)
(6) previous experience indicates that other errors may come up
when the system load is high, and much sooner, i.e. within
seconds of starting the download. The termination of sz does
not always kill the session, I have been able to resume the
download from the download choices page using Kermit
successfully, sometimes.
System load was 3,3,2 when I logged back in, this time.
Just a lot of questions at this stage, but at least
its a little hard evidence. Can anybody volunteer to check the
source? I doubt if I will get a chance today :-(
Won't likely be at the Saturday Session today either (regrets...)
Ed Dyer aa146@chebucto.ns.ca (902) H 826-7496 CCN Assistant Postmaster
http://www.chebucto.ns.ca/~aa146/ W 426-4894 CSuite Technical Workshop
Le Ven, 6 Mar 1998, David Potter écrivait :
> Date: Fri, 6 Mar 1998 08:38:56 -0400 (AST)
> From: David Potter <potter@csuite.ns.ca>
> Reply-To: csuite-tech@chebucto.ns.ca
> To: John Nemeth <jnemeth@cue.bc.ca>,
> csuite-tech <csuite-tech@chebucto.ns.ca>
> Subject: Re: Z-modem
>
>
> Hi John... We've been puzzled by zmodem also...
>
> DavidT's recently re-written the CSuite 'idled' to look for data sent as
> well as data returned... He reports zmodem sending extended data streams
> (perhaps after error checking on shorter streams...). These long
> transmissions seem to have been fooling the old 'idled' into thinking the
> session was idle (and booting users off the system.) Now we look for both
> data in and out during the idle period, but after we warn the user we only
> look for data/keystrokes back from the user. This is not yet in the
> distribution... it needs to be cleaned up a little...
>
> Ed Dyer (our postmaster and zmodem checker outer...) reports that zmodem
> will fail, drop him back to the download screen and that he has been able
> to successfully re-start/complete the download with kermit. DavidT
> hypothesized that, with extended data streams, if zmodem gets an error at
> the end of the transmission it may decide there's nothing to do but
> re-start the whole thing ...while kermit is picking up errors (and
> resending packets) as is goes along...
>
> Lot's of hypothesis here... I'll see if I can track down the author...
> (just something else on my list...)
>
>
> david potter
>
>
>
> On Thu, 5 Mar 1998, John Nemeth wrote:
>
> > We've had a number of complaints about problems with our z-modem
> > file transfer software, which didn't come with CSuite. Can anybody
> > tell me where to get a good version of z-modem that is free? Failing
> > that, what are the recommendations for commercial versions?
> >
>
next message in archive
no next message in thread
previous message in archive
previous message in thread
Index of Subjects