[Discuss] network switch

[adler at stephenadler.com (Stephen Adler)]

On Tue, 2014-07-15 at 17:46 +0000, Edward Ned Harvey (blu) wrote:
> > From: discuss-bounces+blu=nedharvey.com at blu.org [mailto:discuss-
> > bounces+blu=nedharvey.com at blu.org] On Behalf Of Stephen Adler
> > 
> > I was thinking that I may need to get a new switch for a set of
> > computers I have in a lab at work. 
> 
> If you're thinking about using vlan's to separate traffic in a lab, which is potentially bad traffic (like, you're supporting a team of developers who are working on the MAC layer of some new hardware chip for example) I'll recommend:  Don't.   Been there, done that.
> 
> Here's an anecdote:  I worked at a company making wifi chips.  We isolated the lab network with vlan's, so if the engineers screwed anything up, they would only bring down their lab network.  In theory.  Turns out, not in practice.  One incident, their chip started spewing the network with MAC addresses that were all 0's.  That brought down the whole building.  
> 
> 

Thanks, the good news is that its a medical/biology lab. They may be
processing DNA sequences but that shouldn't get routed into the link
layer of the switch... :)

Cheers.