Changing "Class C" Network to CIDR - TCP-IP

This is a discussion on Changing "Class C" Network to CIDR - TCP-IP ; Hi Folks, This is really to check my work, so if anyone has a sec and can see if what I'm planning is cookoo I'd appreciate it (I'm mainly a programmer not a network admin, so I may be way ...

+ Reply to Thread
Results 1 to 6 of 6

Thread: Changing "Class C" Network to CIDR

  1. Changing "Class C" Network to CIDR

    Hi Folks,

    This is really to check my work, so if anyone has a sec and can see if
    what I'm planning is cookoo I'd appreciate it (I'm mainly a programmer
    not a network admin, so I may be way of base here):

    We currently have a basic LAN 192.168.0/24. We expect to grow past 253
    hosts in the next year or two, so want to make our network "bigger"
    and let us do some subnetting in the future for various reasons. So,
    here's what we're thinking: Changing the mask to 255.255.224.0 will
    create a "192.168/19" network that will effectively give us eight
    subnets of 8190 each:

    192.168.0.1-192.168.31.254 ("Main" LAN)
    192.168.32.1-192.168.63.254
    192.168.64.1-192.168.95.254
    192.168.96.1-192.168.127.254
    192.168.128.1-192.168.159.254
    192.168.160.1-192.168.191.254
    192.168.192.1-192.168.223.254
    192.168.224.1-192.168.255.254

    We can then leave the servers routers and other important static IPs
    at the addresses they're at now with no "repointing" of various
    clients and apps to new locations, as long as those clients are in
    that first "main" range of addresses.

    Does that make sense or do I have a lot more reading up on subnetting
    to do? Also, is there a reason NOT to use 192.168 (which I always
    thought of as "Class C territory") in this way?

    Thanks for any input,

    Jeff


  2. Re: Changing "Class C" Network to CIDR

    Jeff wrote:
    > Also, is there a reason NOT to use 192.168 (which I always
    > thought of as "Class C territory") in this way?
    >


    You should be fine. The 192.168 block is 192.168.0.0 through
    192.168.255.255. It contains enough addresses for 256 classful Class-C
    subnets, and that is where people usually assign "Class-C"ish private
    subnets, but the entire 192.168 block is certainly available for your use.

  3. Re: Changing "Class C" Network to CIDR

    On Sat, 26 Jan 2008, in the Usenet newsgroup comp.protocols.tcp-ip, in article
    <6780c475-09ff-4968-91c1-02b8740f48f6@s19g2000prg.googlegroups.com>, Jeff wrote:

    NOTE: Posting from groups.google.com (or some web-forums) dramatically
    reduces the chance of your post being seen. Find a real news server.

    >We currently have a basic LAN 192.168.0/24. We expect to grow past 253
    >hosts in the next year or two, so want to make our network "bigger"


    OK - that is done most easily by changing the mask on ALL systems on
    this network. A /23 (255.255.254.0 or 0xffffe00) gives you a maximum
    of 510 hosts, while a /22 (255.255.252.0 or 0xffffc00) gives you a
    maximum of 1022 hosts. I'd be a bit leery of going any wider on a
    single based on possible traffic density. What you need to look at
    is "who is talking to who" and also the size of your switches.

    >and let us do some subnetting in the future for various reasons.


    Subnetting in the future is of little concern now. It means a router
    to connect between subnets.

    >So, here's what we're thinking: Changing the mask to 255.255.224.0
    >will create a "192.168/19" network that will effectively give us
    >eight subnets of 8190 each:


    Those subnets might get awfully busy with 8000 hosts trying to talk
    all at the same time. But there is nothing that requires you to
    have the subnets in the same general address area. You can just as
    easily have 1024 /22 subnets in the 172.16.0.0/12 range hanging off
    a router, or if you want to go nuts - 16384 /22 subnets in the
    10.0.0.0/8 range.

    >We can then leave the servers routers and other important static IPs
    >at the addresses they're at now with no "repointing" of various
    >clients and apps to new locations, as long as those clients are in
    >that first "main" range of addresses.


    If all you are doing is changing the network mask, then nothing has
    to change except for possibly cosmetic reasons (it looks prettier,
    but otherwise makes absolutely no difference). Hosts on the other
    networks can _probably_ use the same _servers_ but will have to know
    the _local_ gateway address of the router. Ever look at the output
    of the command /sbin/route -n on a *nix box? You'd likely see
    something looking like

    [example ~]$ /sbin/route -n
    Kernel IP routing table
    Destination Gateway Genmask Flags Metric Ref Use Iface
    192.168.0.0 0.0.0.0 255.255.254.0 U 0 0 89948 eth0
    192.168.2.0 192.168.1.6 255.255.254.0 UG 0 0 32165 eth0
    127.0.0.0 0.0.0.0 255.0.0.0 U 0 0 388 lo
    0.0.0.0 192.168.1.248 0.0.0.0 UG 0 0 2673 eth0
    [example ~]$

    Here, there is the "local" LAN that this box is attached to, using
    192.168.1.0/23. There is another subnet - 192.168.2.0/23, and if I
    want to talk to it, I send my packets to the router at 192.168.1.6.
    There is also a route "to the world" (which really means "everywhere
    else") reachable via a router at 192.168.1.248. No mess, no fuss,
    and all blindingly simple.

    >Does that make sense or do I have a lot more reading up on subnetting
    >to do? Also, is there a reason NOT to use 192.168 (which I always
    >thought of as "Class C territory") in this way?


    Only if you have more than 65000 systems give or take. In that case
    you'll have to spill over into 172.16.0.0 - 172.31.255.255 (which
    gives you another million addresses), or 10.0.0.0 - 10.255.255.255
    (which gives you another 16.7 million). Your systems - do as you
    wish. Just don't let those addresses be seen outside your bailiwick.
    Your upstream may be using those addresses for their own nefarious
    purposes, and you don't want to confuse things further.

    Seriously - think about traffic density, as that may be a bigger
    problem than merely making larger subnets.

    Old guy

  4. Re: Changing "Class C" Network to CIDR

    Thanks for all the good advice. Much appreciated.

    Jeff

  5. Re: Changing "Class C" Network to CIDR

    Moe Trin wrote:
    > On Sat, 26 Jan 2008, in the Usenet newsgroup comp.protocols.tcp-ip,
    > in article
    > <6780c475-09ff-4968-91c1-02b8740f48f6@s19g2000prg.googlegroups.com>,
    > Jeff wrote:
    >
    > NOTE: Posting from groups.google.com (or some web-forums) dramatically
    > reduces the chance of your post being seen. Find a real news server.


    While I agree google groups is far from the ideal UseNet posting and
    reading client (especially in it's current form), I dont' think it's
    right to tell anyone what to use. Everyone has personal preferences, he
    may have a vail reason for using google - not all networks will allow
    NNTP connections, like some corporate or school networks.

    I don't like google groups as much as you and others, but I think it's
    wrong to impose such a thing on others without any regard for any
    potential valid reaons for using it.



  6. Re: Changing "Class C" Network to CIDR

    On Mon, 28 Jan 2008, in the Usenet newsgroup comp.protocols.tcp-ip, in article
    <606pmiF1pekosU1@mid.individual.net>, Steve K. wrote:

    >Moe Trin wrote:


    >> in article
    >> <6780c475-09ff-4968-91c1-02b8740f48f6@s19g2000prg.googlegroups.com>,
    >> Jeff wrote:
    >>
    >> NOTE: Posting from groups.google.com (or some web-forums) dramatically
    >> reduces the chance of your post being seen. Find a real news server.

    >
    >While I agree google groups is far from the ideal UseNet posting and
    >reading client (especially in it's current form), I dont' think it's
    >right to tell anyone what to use.


    That note is included by my news reader when I reply to articles
    posted via groups.google.com It's put there to explain why posts
    from google (and others) may not receive responses.

    >Everyone has personal preferences, he may have a vail reason for using
    >google - not all networks will allow NNTP connections, like some
    >corporate or school networks.


    I'm a network admin - I'm aware of those restrictions, which are normally
    placed because of abuse. However, a few seconds searching with any
    search engine will turn up lists of alternative or free news servers
    many of which do have web interfaces.

    >I don't like google groups as much as you and others, but I think it's
    >wrong to impose such a thing on others without any regard for any
    >potential valid reaons for using it.


    I'm imposing nothing on anyone - I'm advising that there may be a limit
    on the number who will see his post and why, while actually responding
    to the content of his post.

    Old guy

+ Reply to Thread