[net.unix-wizards] Making sense of LOGFILE

jcn@aplvax.UUCP (05/02/84)

I'm trying to strip info out of the LOGFILE created by uucp.
Here is a reprint of a section of that file for purposes of discussion:

	uucp milo (3/19-15:44-3150) /usr/spool/uucp/LCK.SQ (CAN'T LOCK)
	uucp milo (3/19-15:44-3150) OK (startup)
	eric milo.p (3/19-15:44-3725) OK (startup)
	eric milo.p (3/19-15:44-3725) REQUEST (S /etc/group /tmp/milogrp eric)
	uucp milo (3/19-15:44-3150) REQUESTED (S /etc/group /tmp/milogrp eric)
	eric milo (3/19-15:44-3150) COPY (SUCCEEDED)
	eric milo.p (3/19-15:45-3725) REQUESTED (CY)
	eric milo.p (3/19-15:45-3725) OK (conversation complete)
	eric milo (3/19-15:45-3150) OK (conversation complete)

I can interpret the first three fields okay, but the last portion
is a bit obscure.  A recent submission to net.sources assumed
that REQUEST always meant that a file was being sent to a remote
system and COPY meant that a file was received from a remote system.
I have no problem with the assumption concerning COPY, but a quick
look at the source code seems to indicate that REQUEST can be attached
to send OR receive.  Is there any guide available which explains what
the various fields of the LOGFILE mean?  The source code is a rather
obscure source of info.

				Thanks,
-- 

					John Noble
					JHU/APL
				...decvax!harpo!seismo!umcp-cs!aplvax!jcn
				...rlgvax!cvl!umcp-cs!aplvax!jcn

honey@down.UUCP (05/04/84)

interpreting your log is complicated by the fact that milo.p and milo
are the same machine; the logs for both sides of the transaction are
merged into the same file.  splitting by pid into 2 logs simplifies
things.

the milo.p log: (you skipped the SUCCEEDED (call to milo) line ...)
	eric milo.p (3/19-15:44-3725) OK (startup)
initial handshake ok.
	eric milo.p (3/19-15:44-3725) REQUEST (S /etc/group /tmp/milogrp eric)
try to send milo.p!/etc/group -> milo!/tmp/milogrp
	eric milo.p (3/19-15:45-3725) REQUESTED (CY)
milo says it worked.
	eric milo.p (3/19-15:45-3725) OK (conversation complete)
hangup handshake ok.

the milo log:
	uucp milo (3/19-15:44-3150) /usr/spool/uucp/LCK.SQ (CAN'T LOCK)
this is an artifact of milo.p calling itself (but doesn't matter since
no one uses private sequence numbers anyway.)
	uucp milo (3/19-15:44-3150) OK (startup)
initial handshake ok.
	uucp milo (3/19-15:44-3150) REQUESTED (S /etc/group /tmp/milogrp eric)
try to receive milo!/tmp/milogrp <- milo.p!/etc/group 
	eric milo (3/19-15:44-3150) COPY (SUCCEEDED)
success.
	eric milo (3/19-15:45-3150) OK (conversation complete)
hangup handshake ok.

concerning your comment
	The source code is a rather obscure source of info.
we will probably never agree on this point.
	peter honeyman