[Info-vax] Routing when using two interfaces/networks.

[Jan-Erik Soderholm]

Bart.Zorn at gmail.com wrote:
> On Jan 17, 3:28 pm, Jan-Erik Soderholm <jan-erik.soderh... at telia.com>
> wrote:
>> Hi.
>> We have some hard-to-identify network problems at
>> a site and I'm now thinking that it might have something
>> to do with how the routing is setup in the VMS system.
>>
>> The VMS system (the "prod" system) has two interfaces :
>>
>> $ tcpip sh inter
>>
>> Interface   IP_Addr         Network mask
>>
>>   LO0        127.0.0.1       255.0.0.0
>>   WE0        193.183.98.2    255.255.255.0
>>   WE1        10.32.137.1     255.255.255.0
>>
>> The 193.183.x.x is the "old" network that will be removed
>> anytime(tm). The 10.32.x.x network is where most of the other
>> stuff (term-servers, PC-clients and so on) are. Note that
>> most other equipment on 10.32.x.x are on other subnets then
>> 10.32.137.x. The only other host on the 10.32.137.x are mainly
>> the other VMS systems. All "user" equipment are on other
>> 10.32.x.x networks.
>>
>> The current routing looks like this :
>>
>> $ tcpip sh route
>>
>>                         DYNAMIC
>>
>> Type           Destination                    Gateway
>>
>> AN    0.0.0.0                         193.183.98.251
>> AN    10.32.137.0/24                  10.32.137.1
>> AH    10.32.137.1                     10.32.137.1
>> AH    127.0.0.1                       127.0.0.1
>> AN    193.183.98.0/24                 193.183.98.2
>> AH    193.183.98.2                    193.183.98.2
>> st
>> $ tcpip sh route/perm
>>
>>                         PERMANENT
>>
>> Type           Destination                    Gateway
>>
>> PN    0.0.0.0                         193.183.98.251
>>
>> The "dev" system has only one interface and looks like this :
>>
>> $ tcpip sho inter
>>
>> Interface   IP_Addr         Network mask
>>
>>   LO0        127.0.0.1       255.0.0.0
>>   WE1        10.32.137.3     255.255.255.0
>>
>> $ tcpip sh rout
>>
>>                               DYNAMIC
>>
>> Type           Destination                   Gateway
>>
>> AN    0.0.0.0                          10.32.137.254
>> AN    10.32.137.0/24                   10.32.137.3
>> AH    10.32.137.3                      10.32.137.3
>> AH    127.0.0.1                        127.0.0.1
>>
>> $ tcpip sh rout/perm
>>
>>                               PERMANENT
>>
>> Type           Destination                   Gateway
>>
>> PN    0.0.0.0                          10.32.137.254
>>
>> The "problems" we are seeing are e.g. :
>>
>> - Troubles ("hangs") when FTP copying from "prod" to "dev".
>> - Intermittent slow access from PC clients.
>>
>> My guess is that the disturbances are due to the fact that all
>> routing goes through the 193.183.98.251 gateway, even between
>> different 10.32.x.x subnets, right ? And that the solution
>> probably would be to simply move the default router from
>> the 193.183.98.251 gateway to the 10.32.137.254 gateway on
>> the system with two networks ?
>>
>> Jan-Erik.
> 
> Two questions here:
> 
> - Does your production systems depend heavily on TCP/IP connections
> (most likely yes) and on which subnet do most of the clients reside?

Yes. There are some older LAT based DECservern that I think are on
both nets/interfaces.

> - Is the 10.32.x.x network subdivided with routers or is it a single
> network like 10.32.0.0/16?

Not, thgere are 5-10 10.32.x.0/8 nets. The VMS boxes are more or
less alone in one of them. All user equipment (termserver and PC's
mainly) are on other 10.32 subnets.


> 
> If you have problems in the current setup with traffic between prod
> and dev, you might have similar problems between prod and it's clients
> when you change the default routing.

Hm, maybe. I'll probably do some tests according to the post from JF
anyway. That is, having routing to both nets setup.



> 
> Bart Zorn