[K12OSN] Tuning LTSP Performance

Terrell Prude' Jr. microman at cmosnetworks.com
Sat Aug 30 03:48:08 UTC 2008


Robert Arkiletian wrote:
> On Fri, Aug 29, 2008 at 8:23 PM, Almquist Burke
> <balmquist at mindfirestudios.com> wrote:
>   
>> ARP poisoning? All they need is access to the network on a machine with raw
>> socket capabilities.
>>     
>
> Okay basically MAC spoofing eth0 on the server. But if they do this
> won't it kill all the X sessions running on the clients? Plus won't
> the switch get confused as to which port to send the data to since it
> will now have 2 ports claiming the same MAC address?
>   

Yeah, it could kill the X sessions, but the switch itself likely won't 
get confused.  It'll simply use the last entry that was made in the CAM 
table, assuming that it has enough CPU to keep its CAM tables updated 
fast enough (two boxes saying, "aa-bb-cc-dd-ee-ff is over here!"  "Oops, 
aa-bb-cc-dd-ee-ff is over *here* now!" and so on).  Remember that 
bridges are not very complicated; the logic's pretty simple.

What *will* get confused is any X11 sessions that are expecting certain 
packets to arrive, and those packets don't, and that's how the sessions 
get degraded, then killed.

--TP




More information about the K12OSN mailing list