[c-nsp] Appletalk (don't laugh) through Cisco switch

[robbie]

A couple of ideas, off the top of my head...

Possible ethertype problems? I know that in the Extreme world, sometimes 
you would have problems if you didn't define ethertypes to be used on a 
specific VLAN, causing it to sort of spaz out if it didn't automagically 
recognize the ethertype field in the frame when it received it (older 
Extremeware, IIRC).

Also, found this: 
http://archive.mac-mgrs.org/archive/2004/20040109183508.txt

Might be pertinent. Have you turned off IGMP snooping on the 3550? In 
12.1(19)EA1c EMI it is turned on by default. Perhaps intercepting frames 
and doing odd, unspeakable things to them?

Some information about your IOS and specific configuration would be helpful.

HTH,
Robbie

Jeff Kell wrote:
> Jeff Kell wrote:
> 
> 
>>"We have researched this problem and discovered that there is a known
>>issue with running AppleTalk through a Managed CISCO Switch."
> 
> 
> To answer some replies already received, and to clarify the original
> situation...
> 
> 
>>Are all the Mac's on the same Vlan on that switch?
> 
> 
> Yes.  It is an old 8Mb 2924-XL actually now (the 3550 was just the 
> uplink), we tried a 2950 with same result, and "downgraded" to the 
> 2924XL to see if it helped).  The ports in question are all on vlan192.
> 
> 
>>Apple computers come up on the network card very fast.. And I found
>>that I had to enable portfast on every port...
> 
> 
> We do that religiously on host ports (portfast).  Configuration of the
> ports is simply:
> 
> 
>>interface FastEthernet0/xx switchport access vlan 192 spanning-tree
>>portfast
> 
> 
> We do have some subnets in the core routing Appletalk, just not to this
> particular building, and have never had issues before.  Been doing this
> long time no problems.
> 
> 
> 
>>Jim is right about the portfast.  That has caused me a lot of grief,
>>especially in labs where netboot is used a lot.
> 
> 
> We have a G5 server netbooting two labs across subnets.  Not afraid of 
> Appletalk :-)
> 
> 
>>Older Novell clients (using IP) did the same thing...portfast worked
>>like a charm.
> 
> 
> We have legacy Novell over IPX, and new Netware Client32 over IP.  No 
> problems there either.
> 
> Which leaves me with the question, what can this application possibly be 
> doing Appletalk-ish that wouldn't work on a plain old layer2 2924XL? 
> And note that their recommended "solution" is to use a non-managed 
> switch (they did not say "hub" which I might have bought into being 
> somehow different).
> 
> Jeff
> _______________________________________________
> cisco-nsp mailing list  cisco-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/
> 
> 


-- 
Cheers,
Robbie