bob@tut.cis.ohio-state.edu (Bob Sutterfield) (06/24/88)
Though we are a news neighbor of Rutgers, we got our four iterations
of the sendsys via our connections to bloom-beacon and husc6, and they
already had long paths (beyond that) on them by the time they got
here.
Because we seem to have been in the (apparently forged) Path: lines of
a lot of the sendsys messages your machines received, many the mail
replies seem to have tried to pass through here. For the past few
days, poor rutgers.rutgers.edu has been on its knees under the load
(likely of all the incoming mail) - its sendmail daemon is even having
trouble answering at the SMTP socket when our sendmail daemon runs the
queue and tries to deliver mail over the Internet.
This means that mail from hither (Ohio State, or likely any other
Internet site) to yon (Rutgers) isn't flowing very well. This means
that, with all your sys files passing through, the mail queue on
tut.cis.ohio-state.edu reached an all-time record of something over
*600* items by late Thursday morning.
Fortunately, Tut's /usr/spool is an Eagle "a" partition, just over
19Mb (/usr/spool/news is an "i" partition, 376Mb on another drive on
another controller), and since most sys files are mercifully short, we
happened not to overflow the space available before discovering and
dealing with the situation.
I asked an assistant (Thanks, Chris - sorry to dump that one on you!)
to rm from the queue the qf and df files of all those that were
sendsys replies. Our mail queue now stands at 55 items, about five
times its normal steady-state level, but at least it's no longer fifty
times. Many of those remaining messages are also awaiting a chance to
connect with Rutgers' SMTP socket - just our normal traffic that's
backed up for a few days. Other mail of yours may be among it, but
your sendsys reply isn't.
Please don't feel disappointed if Bob Webber doesn't get to personally
read your very own sys file, whether he was the one who asked for it
or not. And I don't want to hear anything about what may, in another
situation, legally amount to reading and discarding other people's
private mail as it tried to transit our system. It was self-defense.
Sue the person who forged the sendsys, not me.
--
Bob Sutterfield, Department of Computer and Information Science
The Ohio State University; 2036 Neil Ave. Columbus OH USA 43210-1277
bob@cis.ohio-state.edu or ...!{att,pyramid,killer}!cis.ohio-state.edu!bobwebber@porthos.rutgers.edu (Bob Webber) (06/25/88)
In article <16253@tut.cis.ohio-state.edu>, bob@tut.cis.ohio-state.edu (Bob Sutterfield) writes: > > Please don't feel disappointed if Bob Webber doesn't get to personally > read your very own sys file, whether he was the one who asked for it > or not. And I don't want to hear anything about what may, in another > situation, legally amount to reading and discarding other people's > private mail as it tried to transit our system. It was self-defense. > Sue the person who forged the sendsys, not me. Just in case anyone is unclear about it, I have NO objections to people zapping this stuff (although I am collecting what does get thru just to see what it all says). If you find it necessary to look at the contents of the file to feel justified in zapping it, for pete's sake go ahead and look. What I recieve will eventually be read by automatic scripts. If you want to send me mail, for pete's sake don't send thru rutgers!webber (but rather send to rutgers!athos.rutgers.edu!webber which is the address that always appears in my signature -- as far as I know, no one has ever used the rutgers!webber alias for anything, it exists for historical reasons of backward compatibility of mail paths when machines changed names). Also, don't send from an administrative account (mail from root, guest, nobody, etc., is simply being saved off into a file for later processing). So far [est noon sat], I have recieved 2.8 megabytes of mail (which means that a bit under 1 meg compressed) that I have saved as sendsys replies. Presumably that means rutgers has in three days turned about 9 megabytes of mail over this matter (3 in, 3 out to me, 3 back to sender). When Mel gets back, I will doubtless find out what percentage of the normal rutgers load this is. ---- BOB (webber@athos.rutgers.edu ; rutgers!athos.rutgers.edu!webber) p.s., And I am still quite certain I have never in my life send out any sendsys requests (and I certainly have no plans to do so in the future).