"Steve Winograd [MVP]" <bcma***@mvps.org> wrote in message
news:ner3431slpep7sivhbpen8uif27n7et2u0@4ax.com...
> In article <Ozr#NEgkHHA.3***@TK2MSFTNGP04.phx.gbl>, "Fred Marshall"
> <fmarshallx@remove_the_x.acm.org> wrote:
>>I want to clarify something about setting up routes in Windows XP:
>>
>>If there's a general subnet route:
>>
>>192.168.1.0   255.255.255.0    192.168.1.2
>>
>>(which I understand to mean that anything on subnet 192.168.2.0 goes to
>>192.168.2.1
>
> That route says that traffic for anything on subnet 192.168.1.0 goes
> to 192.168.1.2.
>
> Here's a route that says that anything on subnet 192.168.2.0 goes to
> 192.168.2.1.  Is this what you meant?
>
>   192.168.2.0   255.255.255.0  192.168.2.1
>
>>And, in addition, one needs a specific route within that subnet that's
>>different, then it appears that:
>>
>>192.168.2.33 255.255.255.0     192.168.1.17
>>
>>doesn't work.
>
> That route doesn't work because no IP address can match it.  There's a
> zero in the fourth octet of the subnet mask, but the fourth octet of
> the destination address is 33:
>
>   192.168.2.33 AND 255.255.255.0 = 192.168.2.0
>
>>But
>>
>>192.168.2.33 255.255.255.255     192.168.1.17
>>
>>does work.
>>
>>Why?
>
> That route works because IP address 192.168.2.33 matches it:
>
>   192.168.2.33 AND 255.255.255.255 = 192.168.2.33
>
>>However.... the entry for the last one appears in the routing table after
>>the more general one.  So, it wouldn't be reached it appears. ???
>
> When more than one route can reach a destination address, the route
> with the longest subnet mask is used.  The order in which routes
> appear in the route table is irrelevant.
>

>>>I want to clarify something about setting up routes in Windows XP:
>>>
>>>If there's a general subnet route:
>>>
>>>192.168.1.0   255.255.255.0    192.168.1.2
>>>
>>>(which I understand to mean that anything on subnet 192.168.2.0 goes to
>>>192.168.2.1
>>
>> That route says that traffic for anything on subnet 192.168.1.0 goes
>> to 192.168.1.2.
>>
>> Here's a route that says that anything on subnet 192.168.2.0 goes to
>> 192.168.2.1.  Is this what you meant?
>>
>>   192.168.2.0   255.255.255.0  192.168.2.1
>>
>>>And, in addition, one needs a specific route within that subnet that's
>>>different, then it appears that:
>>>
>>>192.168.2.33 255.255.255.0     192.168.1.17
>>>
>>>doesn't work.
>>
>> That route doesn't work because no IP address can match it.  There's a
>> zero in the fourth octet of the subnet mask, but the fourth octet of
>> the destination address is 33:
>>
>>   192.168.2.33 AND 255.255.255.0 = 192.168.2.0
>>
>>>But
>>>
>>>192.168.2.33 255.255.255.255     192.168.1.17
>>>
>>>does work.
>>>
>>>Why?
>>
>> That route works because IP address 192.168.2.33 matches it:
>>
>>   192.168.2.33 AND 255.255.255.255 = 192.168.2.33
>>
>>>However.... the entry for the last one appears in the routing table after
>>>the more general one.  So, it wouldn't be reached it appears. ???
>>
>> When more than one route can reach a destination address, the route
>> with the longest subnet mask is used.  The order in which routes
>> appear in the route table is irrelevant.
>
>Steve,
>
>Thanks again!  That's very useful to know.  Where might one find these kinds
>of "rules"?  I've Googled but nothing seemed to pop out that did it.