[NTLUG:Discuss] ntlug.org back from the dead again

[Chris Cox]

Burton Strauss wrote:
> Even hard-coded, switches are supposed to advertise themselves as their hard
> coded attributes and it SHOULD work. But as you point out, old gear often
> mis-implemented that.
> 

AFAIK, that is not true.  That's my understanding of the protocol though.
I know that here, with expensive modern (today's model) Cisco switches,
that if the switch port is configured 100 FD and the client is auto
negotiate, that we get 100 Half with errors.


> As for why 10 - it's an effective albeit crude form of bandwidth limitation.
> 
> -----Burton 
> 
> -----Original Message-----
> From: discuss-bounces at ntlug.org [mailto:discuss-bounces at ntlug.org] On Behalf
> Of Chris Cox
> Sent: Tuesday, July 05, 2005 4:17 PM
> To: NTLUG Discussion List
> Subject: Re: [NTLUG:Discuss] ntlug.org back from the dead again
> 
> joseph beasley wrote:
>>We normally force servers to 100/full and workstations to auto.  
>>Unless it is my desktop of course.  Then it's 100/full also.
> 
> By definition if one side is auto and the other is fixed, you'll go end up
> with half.... however one side will probably believe it is full and the
> other half... and many switches will dumb it all the way down to 10baseT
> half because of all of the errors.
> 
> Speak up network guys!
> 
> My understanding from my own experience:
> 
> Auto+Auto=works
> Auto+Hard coded=no workie (theplanet.com)
> Hard coded+Hard coded=works if both set the same
>    (this is what ntlug.org is doing now that
>     I know they hard code their ports)
> 
> <snip />
> 
> 
> _______________________________________________
> https://ntlug.org/mailman/listinfo/discuss
> 
>