begin  quoting kelsey hudson as of Tue, Jul 25, 2006 at 01:56:21PM -0700:
> Stewart Stremler wrote:
[snip]
> >At contract[-3], the network would collapse every day between 2 and 3 pm,
> >for at least half-an-hour, up to an hour-and-a-half.
> 
> Sounds like you either need to:
> a) Fire your network admins and get someone competent

Yah, right. I was on a small project. It was a lucky thing I wasn't
forced on to NMCI.

> b) Find the piece of hardware responsible for this and eliminate/replace it.

Not my job. I have neither the power, authority, or funding to do such
a thing.
 
(I did discover that someone on the subnet was blasting NetBIOS packets
and saturating the network on a regular basis.  That took care of the
normal 10AM network slowdown. We threatened to cut the guy's network
cable.)

> A good corporate network should *not* go down unless there's 
> catastrophic hardware failure in your core -- and that's what 24/7/4hr 
> service contracts are for. It's safe to 8/5/NBD stuff that isn't mission 
> critical (like a distribution switch or something), but you should have 
> a spare anyways.

Well, yes.

> I just don't understand how anyone can run a network that constantly 
> goes down. It's like they're running embedded windows on their switches 
> or something :D

Could be.

If the network goes down on a predictable basis, you can adapt, unless
your tools are *entirely* network based.

-- 
_ |\_
 \|


-- 
[email protected]
http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list

Reply via email to