Re: Hysterical first technical alert from US-CERT

[Valdis.Kletnieks () vt edu]

On Wed, 04 Feb 2004 09:41:39 EST, Larry Seltzer said:

> The advisory specifically says that MyDoom.B is spreading rapidly, and that was never
> the case. It didn't say that it *could* spread rapidly. Maybe you think misinforming in
> order to induce caution is a good idea, but I expect nothing but the truth from an
> agency like this.

And I posted a heads-up to our local staff about Dumaru a lot quicker than
I did for MyDoom, because from where *I* was, I saw a *huge* initial spike
of Dumaru.  If I had waited, I would have realized that Dumaru had fizzled.
On the other hand, if I had waited that long and it took off like MyDoom,
we'd have been screwed.

As I said - would you rather they delayed 12 or 18 hours to identify
*for sure* how fast it was spreading?  Read Nick Weaver's work on
Warhol Worms at http://www.cs.berkeley.edu/~nweaver/warhol.html and then
ask yourself how much time they should wait and verify before releasing.

Unless you have *proof* that they already *knew* it was a snoozer when
they hit send, or you have *specific* recommendations on how they can
do better, let it slide.

Or alternatively, what would *YOU* do if your boss at Ziff Davis told you that
there were cases where your article *had* to be on the web server *within an
hour* of you getting the first hint of the story, or real damage might happen?
Oh, and you don't know which stories those are, and which ones you can afford
to wait 2 or 3 hours and do follow-ups on first. Oh, and Ziff Davis also said
that if you screwed up and got a fact wrong, you'd hear about it from all your
readers.

If you got a lead that a massive DDoS was coming in 90 minutes, what would you
do?

Attachment: _bin
Description: