[rescue] Re: ethernet and switches,

[Dave McGuire]

On Apr 23, 2004, at 4:29 AM, Robert Rose wrote:
> Cisco's FDDI implementation wasn't that good in the mid-90's.  At 
> $WORK-7 we had a core of 3 AGS+ connected with an FDDI ring, and 
> bought a few sbus FDDI cards for our new SparcServer 1000's to join 
> them to the ring.  We used to shut down all hosts every Sunday morning 
> to single user mode, ran backups to tape and reboot, so the FDDI ring 
> would wrap while the boxes rebooted.
>
> "Yeah, so what"? you say, well when two boxes adjacent on the ring 
> happened to be rebooting at the same time, the damn Ciscos saw two 
> failures in the ring and shut the whole ring down! Management's 
> solution was to ditch the whole FDDI setup and go fast ethernet.  I 
> couldn't even to convince them to keep an FDDI ring just for the 
> 1000's as a dedicated backup/NFS network, despite the only problem 
> that we ever had was when the Cisco routers were in the ring.

   Yeah, they had some bugs early-on.  That particular one bit us too, 
but it was fixed pretty quickly.  Well, where "quickly" means "shortly 
after we experienced it"...I don't know how long it was actually in 
there.  It sure did suck for a while there, though.

          -Dave

--
Dave McGuire          "PC users only know two 'solutions'...
Cape Coral, FL          reboot and upgrade."    -Jonathan Patschke