[Lias] RH 7.1 control panel

[Les Bell]

Robert McLeay <ches at perlboy.org> wrote:

>>
Selling you a copy?

Ain't it free as in beer *and* speech? ;)
<<

Sure is, but a little sales spiel doesn't hurt free software!

>>
Seriously, though, I wonder how accurate those TCP sequence numbers
relate to uptime. I know that NMap always gives out the incorrect uptime
when scanning the computer I'm sitting at.
<<

Dunno. I think that in theory, TCP initial sequence numbers are supposed to
be pseudo-random. Real-world is a different matter, though.

>>
38.7 day average uptime too :)
<<

Don't read too much into it; I've applied various kernel patches for
security, and at least once, recently, when the Domino server threads
froze, the quickest way to fix it was just to bounce the server. So 38.7
days is probably about right, and ain't too bad considering the workload on
the box.

I used to be quite manic about uptimes - the Domino server got up to 160
days or so at one point, and I was *so* chuffed about that - but I got
caught out badly when a glibc patch I installed wasn't picked up until I
bounced the server, something like six weeks later, and Domino linked with
the new glibc for the first time and promptly fell over. I wasted so much
time backing out other patches that now I'm a lot better organised in my
"configuration management" and will not wince at rebooting the server just
so I can be *sure* that all the patches are still flying in formation.
Better to sleep soundly than suffer uptime hubris. ;)

Best,

--- Les Bell, CISSP
[http://www.lesbell.com.au]