I had dinner in Berkeley the evening of the Morris Worm at Joshu-Ya - the "Kabuki
West" dinner group that Russell Brand started when he moved west from MIT, with some
help from me. Unusually, I went directly bed when I got home to Mountain View instead of
reading E-mail on
apple.com before crashing out. Many of my dinner companions went back to
the eXperimental Computing Facility (XCF - for undergrads) in Cory Hall on the UCB campus,
found their facilities under attack, and coordinated with a team at MIT to perform
analysis. I remember that Dave Pare put the binary analysis skills he'd acquired in
decompiling psl's empire game to good use in analyzing the worm.
I found out the next morning that
apple.com was off the Internet (CSNET had shut off the
X25NET), and that it (a VAX-11/780 running 4.3 BSD UNIX; we upgraded to an 8650 not much
later) had been successfully attacked 17 times overnight ... but that our X25NET
connection (IP over X.25 at 9600 baud) had been so flakey that the worm hadn't
managed to successfully download its second part and start it. I shut off the finger TCP
service, checked to make sure our sendmail(8) didn't have the "debug mode
feature" that the worm exploited, and told CSNET to turn us back on.
Erik Fair, formerly {post,host}master(a)apple.com