pusateri@duke.cs.duke.edu (Thomas J. Pusateri) (10/22/90)
I am having trouble establishing two way communication with HDB uucp.
If any one can suggest any ideas, I would greatly appreciate it them.
Basically, I have a ATT 3b1 3.51m running HDB uucp from THE STORE
connected to a SUN workstation SUNOS 4.1 running HDB uucp. I can
send files from the unix-pc to the sun without any problem but I can
never send files from the sun to the unix-pc. It doesn't matter who
initiates the connection, if the file is going from the unix-pc to
the sun it works.
However, if it is going from the sun to the unix-pc, I always get errors.
The error in the /usr/spool/uucp/.Admin/error looks like this
ASSERT ERROR (uucico) pid: 316 (10/17-11:39:48) PKXSTART ret (0)
[FILE: pk1.c, LINE: 331]
and the output in the audit file looks like the following. Anyone have
any good clues out there?
Thanks for any help
Tom Pusateri
Duke University
---------------------- audit file excerpt -----------------------------
imsg >^M^J^PShere=willow^@Login Successful: System=willow
omsg "Snbsr -x9"
imsg >^PROK^@msg-ROK
Rmtname willow, Role MASTER, Ifn - 7, Loginuser - root
rmesg - 'P' imsg >^PPge^@got Pge
wmesg 'U'g
omsg "Ug"
send CNTL, INITA, val 7
recv CNTL, INITA, val 3
send CNTL, INITB, val 1
state - [INIT code a] (01)
recv CNTL, INITB, val 1
send CNTL, INITC, val 7
state - [INIT code a]&[INIT code b] (03)
recv CNTL, INITC, val 3
state - [O.K.] (010)
Proto started g
*** TOP *** - role=1, setline - X
gtwvec: dir /var/spool/uucp/willow
insert(C.willowN7577) insert C.willowN7577 at 0
return - 8
Wfile - /var/spool/uucp/willow/C.willowN7577,Jobid = willowN7577
Request: nbsr!/etc/group --> willow!/usr/spool/uucppublic/group (root)
wrktype - S
wmesg 'S' /etc/group /usr/spool/uucppublic/group root -dc D.0 644 root
send DATA, loc 1, rem 0
rmesg - 'S' recv CNTL, RR, val 1
state - [O.K.] (010)
send CNTL, RR, val 1
got SY
PROCESS: msg - SY
SNDFILE:
send DATA, loc 2, rem 1
send DATA, loc 3, rem 1
send SHORT, loc 4, rem 1
pkcget: alarm 1
Alarm while looking for SYN -- request RXMIT
send DATA, loc 2, rem 1
pkcget: alarm 2
Alarm while looking for SYN -- request RXMIT
send DATA, loc 2, rem 1
pkcget: alarm 3
Alarm while looking for SYN -- request RXMIT
send DATA, loc 2, rem 1
pkcget: alarm 4
Alarm while looking for SYN -- request RXMIT
send DATA, loc 2, rem 1
pkcget: alarm 5
Alarm while looking for SYN -- request RXMIT
send DATA, loc 2, rem 1
pkcget: alarm 6
Alarm while looking for SYN -- request RXMIT
send DATA, loc 2, rem 1
pkcget: alarm 7
Alarm while looking for SYN -- request RXMIT
send DATA, loc 2, rem 1
pkcget: alarm 8
Alarm while looking for SYN -- request RXMIT
send DATA, loc 2, rem 1
pkcget: alarm 9
Alarm while looking for SYN -- request RXMIT
send DATA, loc 2, rem 1
pkcget: alarm 10
Alarm while looking for SYN -- request RXMIT
send DATA, loc 2, rem 1
pkgetpack failed after 10 tries
send DATA, loc 3, rem 1
send SHORT, loc 4, rem 1
send CNTL, CLOSE, val 0
send CNTL, CLOSE, val 0
cntrl - -1
omsg "OOOOOO"
send OO 0,omsg "OOOOOO"
imsg >^POOOOOO^@exit code -1
Conversation Complete: Status FAILED
TM_cnt: 0upl@gumby.cs.wisc.edu (Undergrad Projects Lab) (10/26/90)
In article <656603007@macbeth.cs.duke.edu> pusateri@duke.cs.duke.edu (Thomas J. Pusateri) writes: [ ... ] >Basically, I have a ATT 3b1 3.51m running HDB uucp from THE STORE >connected to a SUN workstation SUNOS 4.1 running HDB uucp. I can >send files from the unix-pc to the sun without any problem but I can >never send files from the sun to the unix-pc. It doesn't matter who >initiates the connection, if the file is going from the unix-pc to >the sun it works. > >However, if it is going from the sun to the unix-pc, I always get errors. >The error in the /usr/spool/uucp/.Admin/error looks like this >ASSERT ERROR (uucico) pid: 316 (10/17-11:39:48) PKXSTART ret (0) >[FILE: pk1.c, LINE: 331] Oh, my goodness, does this look familiar. I have yet to figure out what exactly is happening. Here are some facts, though. I have a unixpc running HDB retrieved from the store. I have 3 sites with which I regularly communicate via uucp. Only one of these three ever exhibits this problem. It is a site running 286 SCO Xenix. (One of the other sites is also a unixpc and the third is a 386 running ESIX - Everett's unix) My problem seems less deterministic than yours, though. This thing only happens when transferring really big files (>100k) It seems like the other site just decides to stop transmitting. Watching my modem lights confirms that. The other side just stops. Since this thing only happens intermittantly, I just attributed it to the other guy's uucp, and left it be. However, now it seems that it may very well be a bug with our implementation of HDB. If anyone has any tips on how to avoid this, please include me as a recipient of that info. As always, thanks, - sparkie P.S. PLEASE PLEASE PLEASE do not reply to the account from which this has been posted. Instead, reply to: harier!sparkie@cs.wisc.edu Thanks.