On Tue, 2002-10-01 at 16:04, Pavel Kankovsky wrote:
> Gotcha. I myself wondered several times why check_ports.nasl (and its
> buddy port_shell_execution.nasl) spends so much time burning CPU cycles
> without any network activity (*). Sometimes, not always -- and it might
> correspond to the cases when I ran Nessus with a large set of scanned
> ports (maybe even 1-65535).
I have had lockups with both full and partial port ranges. The first I
saw was when the portlist was 1-65535. The second was when the portlist
was "all reserved + nmap services".
OK, wait, just thought of something. Race condition??? In both cases
where I had lockups, two or more machines were being scanned
simultaneously. I've never had a lockup scanning just a single
machine. So could it be that:
i = i + 1;
port = scanner_get_port(i);
The port number list's internal (linked list?) pointer is being updated
by each successive thread, with the result that the list wraps around
and thus triggers an endless loop???
Kris
-
[EMAIL PROTECTED]: general discussions about Nessus.
* To unsubscribe, send a mail to [EMAIL PROTECTED] with
"unsubscribe nessus" in the body.
