Choosing IP Addresses

Emissary_of_PainEmissary_of_Pain Member Posts: 23 ■□□□□□□□□□
Hey all ...

I have a strange question ...


When practicing subnetting or if you have to set up a network from scratch for a company (private network like departments within a company) ... How do you choose a network ...

Do you stick to a certain class C network or do you pick your favourite class C and and have fun from there ...


What I mean is that class C is 192 and above in the first octet and you need to have the first 3 octets reserved (255.255.255.*)

With this in mind ... how do you choose between saying 192 or 216 ... etc ...

(really hope this makes sense)


All information would be greatly appreciated.
Regards.

Comments

  • krjaykrjay Member Posts: 290
    Well I choose 192.168 because it's a reserved private address.

    More specifically though: Usually I've found the questions I'm practicing for CCENT to give you the network and ask you to subnet it such as this:

    "You are designing a subnet mask for the 172.17.0.0 network. You want 110 subnets with up to 300 hosts on each subnet. Which mask do you choose?"
    2014 Certification Goals: 70-410 [ ] CCNA:S [ ] Linux+ [ ]
  • Emissary_of_PainEmissary_of_Pain Member Posts: 23 ■□□□□□□□□□
    What about the 2nd and 3rd octets ? ...

    Isn't anything between 192 and 223 considered reserved ?
  • networker050184networker050184 Mod Posts: 11,962 Mod
    If you want to use anything besides RFC-1918 space you need to have it allocated from your RIR, which is AfriNIC in Africa.

    Most likely you will be using RFC-1918 IPv4 space within your organization though so you can pretty much pick what ever you want in that range. How you pick it really depends on how your network is going to be set up. How many locations and their size, VLANs used etc.

    The RFC-1918 space is:

    10/8
    172.16/12
    192.168/16
    An expert is a man who has made all the mistakes which can be made.
  • Emissary_of_PainEmissary_of_Pain Member Posts: 23 ■□□□□□□□□□
    Ok so lets say you need to set up a network from scratch for a company ...


    They are going to have 6 Departments each with multiple hosts in each department ... how do you choose which IP to use (lets stick to class C)


    Sorry if this all sounds silly but I am used to having the network already in existence and just working with the already determined IP range
  • networker050184networker050184 Mod Posts: 11,962 Mod
    How big are these six departments?

    Personally I'd start like this.

    10.1/16 to site 1
    10.2/16 to site 2
    ...
    ...
    10.255/16 to site 255

    Now you have a /16 for each site to work with. Then you can start breaking your subnets out of that /16 depending on needs of host per subnet. Say you have three VLANs at site 1 all needing a /24.

    VLAN1 - 10.1.0.0/24
    VLAN2 - 10.1.1.0/24
    VLAN3 - 10.1.2.0/24
    An expert is a man who has made all the mistakes which can be made.
  • krjaykrjay Member Posts: 290
    Ok so lets say you need to set up a network from scratch for a company ...


    They are going to have 6 Departments each with multiple hosts in each department ... how do you choose which IP to use (lets stick to class C)


    Sorry if this all sounds silly but I am used to having the network already in existence and just working with the already determined IP range


    You should probably know a rough estimate of how many hosts per subnet before you are dead set on a class C
    2014 Certification Goals: 70-410 [ ] CCNA:S [ ] Linux+ [ ]
  • sratakhinsratakhin Member Posts: 818
    It's better to start designing a new network in 10.x.x.x range, unless you are absolutely sure that it will never need more hosts per subnet than a class C network provides.
  • Emissary_of_PainEmissary_of_Pain Member Posts: 23 ■□□□□□□□□□
    Ah, that makes sense ... thank you ! ...


    The way I am practicing (in packet tracer) is that my GF or who ever gives me a random company name and the departments with hosts in each ...


    Example:

    Metal Car Spares

    6 Departments :
    Sales -2 networks in sales of 20 hosts and 7 hosts
    HR - 15 hosts
    Data Capture - 8 Hosts
    Accounts - 4 Hosts
    Reception - 2 hosts
    Cashiers - 3 hosts

    1 centralized Mail/File server with all departments having Internet Access


    Each department must have expansion room to double in size
  • d6bmgd6bmg Member Posts: 242 ■■■□□□□□□□
    Ah, that makes sense ... thank you ! ...


    The way I am practicing (in packet tracer) is that my GF or who ever gives me a random company name and the departments with hosts in each ...


    Example:

    Metal Car Spares

    6 Departments :
    Sales -2 networks in sales of 20 hosts and 7 hosts
    HR - 15 hosts
    Data Capture - 8 Hosts
    Accounts - 4 Hosts
    Reception - 2 hosts
    Cashiers - 3 hosts

    1 centralized Mail/File server with all departments having Internet Access


    Each department must have expansion room to double in size

    Max 28 hosts per dept with the scope to double it?
    Using the following convention would work:
    192.168.1.0/26
    192.168.1.63/26
    192.168.1.127/26
    192.168.1.191/26
    192.168.2.0/26
    192.168.2.63/26

    will work with basic VLSM. Or you can optimize it for every network out there.
    Or you can use any address space from RFC-1918
    [ ]CCDA; [ ] CCNA Security
  • sratakhinsratakhin Member Posts: 818
    That's not VLSM :) What's the point of giving the cashiers 'department' 62 hosts if they only need 3?
  • Emissary_of_PainEmissary_of_Pain Member Posts: 23 ■□□□□□□□□□
    I agree ... that is a massive waste of host space ... I know that you should plan for growth but that is a bit excessive considering I was told to allow for double growth ...

    These are the subnets I landed up going with ...

    Network 1 : 192.168.5.0 - .31/27
    : 192.168.5.32 - .63/27
    Network 2 : 192.168.5.64 - .95/27
    Network 3 : 192.168.5.96 - .127/27
    Network 4 : 192.168.5.128 - .143/28
    Network 5 : 192.168.5.144 - .147/30
    Network 6 : 192.168.5.148 - .155/29

    Net 7 : 192.168.5.156 - .159/30
    Net 8 : 192.168.5.160 - .163/30
    Net 9 : 192.168.5.164 - .167/30
    Net 10 : 192.168.5.168 - .171/30
    Net 11 : 192.168.5.172 - .175/30
    Net 12 : 192.168.5.176 - .179/30
  • Emissary_of_PainEmissary_of_Pain Member Posts: 23 ■□□□□□□□□□
    Random Question ... busy keying in the IP addresses into Packet Tracer and I ran into a very strange problem ...

    Network 6 - 192.168.5.148 - 192.168.5.155/29

    Now to my understanding I have : .149 for the fa0/0 on the router ... 1 host is .150 and it is fine

    BUT

    When I try key in the next host as either .151 or .152 it says that the IP is invalid for the subnet range ...


    WHY ?!?!?!?! ... ... .153 works perfectly so why wont 151 or 152 ?
  • DiggsDiggs Member Posts: 97 ■■□□□□□□□□
    A /29 (255.255.255.24icon_cool.gif subnet mask give you a block of 8 IP addresses, one of which is used for broadcasts and the other the network ID.

    192.168.5.0
    192.168.5.8
    192.168.5.16
    ...
  • ptilsenptilsen Member Posts: 2,835 ■■■■■■■■■■
    Emissary, the part you are missing about VLSM is that it's only "variable" to the extent that you can define lengths. You cannot arbitrarily set start points of subnets based on the end point of the previous subnet without regard to length. No matter how your subnets are configured, 192.168.5.153/29 is interpreted as being a host in subnet 198.168.5.152/29, with .152 being the node address and .151 being the broadcast address of the previous subnet (if it's configured properly).

    You cannot have network subnets like this:
    1. /27
    2. /28
    3. /30
    4. /29

    The interpretation of the subnet is set in stone by the mask. You don't get to define start points arbitrarily. If you use 192.168.5.13/29, for example, it will be interpreted as 192.168.5.8/29.

    So, in your network, you have a problem between network 5 and 6. Network 4, a /28, can be followed by two /29s, four /30s, etc. or by one /29, then two /30s. It cannot be followed by a /30 and then a /29 at the IP after the end of the /30. Either you leave a /30 unused between Network 5 and 6, or you put your larger subnet first.
    Working B.S., Computer Science
    Complete: 55/120 credits SPAN 201, LIT 100, ETHS 200, AP Lang, MATH 120, WRIT 231, ICS 140, MATH 215, ECON 202, ECON 201, ICS 141, MATH 210, LING 111, ICS 240
    In progress: CLEP US GOV,
    Next up: MATH 211, ECON 352, ICS 340
  • Emissary_of_PainEmissary_of_Pain Member Posts: 23 ■□□□□□□□□□
    So would I just "abandon" Network 5's IP that I assigned and create a Network "13" with a /30 range ? ... would that work then ? ...


    Is it conflicting because network 5 is already on the network ? ...


    (see attached image for network layout using packet tracer)
  • d6bmgd6bmg Member Posts: 242 ■■■□□□□□□□
    sratakhin wrote: »
    That's not VLSM :) What's the point of giving the cashiers 'department' 62 hosts if they only need 3?

    I didn't go for VLSM, as I think OP is preparing for CCNET and VLSM isn't in CCNET syll.
    Or I have made a wrong assumption..icon_sad.gif
    [ ]CCDA; [ ] CCNA Security
  • Emissary_of_PainEmissary_of_Pain Member Posts: 23 ■□□□□□□□□□
    I am preparing for my CCNA actually ...
  • pdosevpdosev Registered Users Posts: 3 ■□□□□□□□□□
    Hi
    The first thing you need to do is to start with the largest to the smallest network. In your case the subnets are:

    hosts network address
    40 192.168.0.0 -192.168.0.63 / 26
    30 192.168.0.64 – 192.168.0.95 /27
    16 192.168.0.96 – 192.168.0.127 / 27
    14 192.168.0.128 – 192.168.0.143 / 28
    8 192.168.0.144 - 192.168.0.159 / 28
    6 192.168.0.160 - 192.168.0.167 / 29
    4 192.168.0.168 - 192.168.0.175 / 29

    But you can't use 192.168.0.40 - 192.168.0.56 /28 because you have 4 bits for host that means each subnet is having 16 spare addresses(14 host space+1 broadcast and network adr) so in this case they are members of either 192.168.0.32 or 192.168.0.48 /28 networks.
    Hope I've helped :)
Sign In or Register to comment.