Q, regarding VLSM and host addresses

Howdy,

If they are less than a decade old they should be able to do VLSM. But since these are security related you might consider sticking them on their own VLAN with their own subnet so you can control access to all of them with one access list.

Matt Sibley Crosswire Labs Real Gear for Real Exams

In general, Hosts (computers, etc.) are not aware of (do not understand) subnetting. They understand only IP address. Subnetting is something that people (administraters) use in order to save public IP addresses.

The Dude

The question was whether the cameras need to be assigned glassful addresses or not. By your statements one could assume you're saying 'yes they do' because unless they are 'aware' of the ability to subnet ie. apply a VLSM then they may not allow, for example, a /25 bit mask to be applied when given an ip address of 195.2.2.2 (a 'C-class' address).

The fact a system allows an classless mask to be applied implies the system supports subnetting. In reality the answer to this question can only be assumed and confirmed only by referring to manufacturers documents or getting hold of a camera and trying.

BernieM

Do you think you're really in a position to make that statement? I mean, do you really know what you're talking about?

Didn't you ask in a previous post ... "Can someone please tell me why don't we put subnetmasks for gateways?"

BernieM

Hey,

Thanks for trying to help, but this is not quite right. Hosts for the most part do understand VLSM. It is just part of the protocol stack code and is in any OS that is current enough to be usefull.

Try it yourself. "netstat /r" on windows or netstat -rn on unix, and you get a picture of the full routing table of your pc with subnet masks. renumber your interface from the control panel or ifconfig to a narrow subnet. Now ping the classfull network outside of your subnet. You should get a host unreachable.

-Psy

Routing tables are created by the routing protocols ....

The Dude

I answer for things that I know, and I ask for things I do not know. That's what the newsgroup is for. Apparently my statement is correct, since the only comment of yours was: "do you really know what you're talking about?" In this case, yes! Hosts understand only 0s and 1s. People implement subnetting through their instructions (software, programs, etc.) on the right hardware platforms. Someone else answered regarding the hardware before me. I thought I would help a little bit with my statement.

The Dude

Uhm, they are not just created by routing protocols. Routing tables are created in a router via static, directly connected, arp'ing AND routing protocols.

You have the wrong attitude. You flame people that try to help you when you don't understand what they're saying. You say they're wrong and you're right when you don't even know what they've told you.

What do you mean 'my only comment'? Didn't you get this post ...

The question was whether the cameras need to be assigned glassful addresses or not. By your statements one could assume you're saying 'yes they do' because unless they are 'aware' of the ability to subnet ie. apply a VLSM then they may not allow, for example, a /25 bit mask to be applied when given an ip address of 195.2.2.2 (a 'C-class' address).

The fact a system allows an classless mask to be applied implies the system supports subnetting. In reality the answer to this question can only be assumed and confirmed only by referring to manufacturers documents or getting hold of a camera and trying.

BernieM

a host is a thing with an ip. or perhaps more generally, a network address. routers are hosts, I have issues with that definition. - 2 directly connected comps technically don't need addresses. are they hosts? I guess so - does the term host apply to any network, any network protocol? But the key point in response to you is that IP routers are IP hosts. With functionaltiy besides IP host functionality

Now i've addressed that, i'll go off on a huge tangent.. (related to some qs you've asked in the thread)

All hosts do what some might call "basic routing", But I don't think it is called routing, I don't think it has a name. It takes the ip it wants to send. looks is it local, is it remote. puts the right MAC in. sends it out. I guess one could call it the IP host function. Or Being an IP host .

The factor that makes a host "act like" an "RFC gateway " (i.e. act like a gatewy according to RFC guidelines) - multihomed (and so takes or can take a packet from another machine and route it to another machine) e.g. comp A is connected to Comp B which is connected to the internet. So compB is multihomed, it's a gateway.

The factor that makes a host a gateway - connects 2 networks (sits between 2 routers?)

A(The?) factor that makes a host an "RFC router" - routing table

-make routing decisions? Yet the term router replaces the term for gateway.- in RFCs And so by that reasoning, we have the requirements for gateways

+requirements for Routers.. Though really, the requirements for routers obsoletes requirements for gateways.

win xp host, a plain old host.

is it a gateway? is it a router? (by RFC standards)

can be answered by 2 questions (does it sit between 2 routers? is it multihomed? does it have a routing table?

does it sit between 2 routers? no

is it multihomed? it could be set up to be, but mine isn't. the facility is there though. If it were multihomed, it'd be "acting as a gateway". But it's not even doing that.

does it have a routing table? it has a routing table. which it uses to do the basic IP host function of sending its own packets out. (it turns sending its own packet into receiving a packet from itself and routing it). A funny use for a routing table. Not really routing! A router should be routing packets from another host.

does it make routing decisions? packets can be sent to 127.0.0.1 or

192.168.1.2 , depending on destination, so that is a routing decision. Though 127.0.0.1 is a logical interface and any packet sent there is not really "routed" - it's not sent out 'elsewhere'!

So, it's inconclusive. RFCs are crap!! But where else do we get definitions?

Host may know about VLSM if the netmask is different from the classful boundaries.

Routing tables are created by software. Routing protocols are is the "language" the software uses to communicate with other routers.

Indeed, I remember a SUN workstation that could only accept clasfull netmasks.

You'll find that some appliance like devices (with some type of an embedded OS) are not capable of using VLSM. Some even depend sorely on proxy-arp.