1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

How big a topic is IP addressing/routing?

Discussion in 'Routing & Switching' started by zr79, Mar 1, 2009.

  1. zr79

    zr79 Byte Poster

    199
    1
    17
    Routing is basically forwarding packets between 2 different subnets so >>192.168.0.1 (255.255.255.0) to 192.168.1.1(255.255.0.0) i think i got that right. ok.

    So a router has to have a routing table before it can forward packets onto the next hop the routing table can be either dynamic or static.??

    Ok so i still haven't clicked with IP addressing and routing, i am basically trying to route packets through 2 cisco routers that are directly joined via ethernet cable. One router is connected to a modem and so to the WAN etc.. and the other router is connected to the first router through an ethernet port with a PC at the end of the second router, i want the PC to have both internet connectivity and beable to connect to any other PC LAN connectivity for any other PC that could be connected to either router..

    I am really stuck on the routing tables and ip addressing. I feel if i can get this sussed i will be 10 steps further on.

    [​IMG]

    Can i have this network here??

    [​IMG]
     
    Certifications: A+
  2. Simon-MCT

    Simon-MCT Bit Poster

    18
    3
    27
    Hi Zr,

    To answer a simple question, yes you can have networks configured as per your second diagram.

    To answer the even more simple question "How big a topic is it?" the answer is "pretty big!"

    Some of the basics: A node on the network, be it a server, desktop or router, typically needs 3 pieces of information to be able to use IP properly; 1) An IP address, 2) a subnet mask and 3) a default gateway

    The node or device only ever knows & uses it's own subnet mask - that realisation was quite a factor in my own understanding of IP.

    Let me ask you a few questions to ascertain your level of knowledge, ok?

    In your first diagram, how many subnets are there, or rather how many subnets do there need to be? Can you come up with an IP addressing scheme for it?

    Same questions for your second diagram.

    Post your answers and we'll take it from there, ok?
     
    Certifications: MCSE:Sec;MCITP; MCTS; MCT; A+,Sec+
    WIP: CCNA
  3. zr79

    zr79 Byte Poster

    199
    1
    17
    delete
     
    Certifications: A+
  4. zr79

    zr79 Byte Poster

    199
    1
    17
    I assume there should be 2 subnets. Suppose the first router was on office floor one and router 2 on office floor 2, both routers are directly joined cabled via ethernet or so (i know you would probably use a switch + router but just for this example i want 2 routers so i can undertand how routers route basically) i am not intersted in using a switch + router here..


    This way this is how i would have thought you would ip address it, now i can ping each router closest the pc from the pc, and the 2 routers in the middle can directly ping each other but i can't ping from pc to pc, i assume becuase i haven't added routing tables, (note that my knowldge of routing tables is not much at all) so,

    1) Is this the right way to address the IPs in the first place
    2) Why can't i ping PC from PC


    [​IMG]
     
    Certifications: A+
  5. zebulebu

    zebulebu Terabyte Poster

    3,748
    330
    187
    You're not doing badly here. In the real world, the link between the two routers wouldn't waste an entire class C subnet - it would be a /30 mask with two usable addresses and be called a point to point link, but apart from that you're pretty much there.

    Regarding routing tables, the easiest way to add routes to a small setup like this would be to add a static routing entry on each of the routers so that they 'know' about the other networks and how to find them. Haver a think about how you would go about doing that and see if you can implement it in packet tracer (hint - you'll find a way of doing this in PT without using the CLI - if you use this, check the CLI afterward so you can see the mode you need to be in and the syntax you need to use. Once you're fine with static routes, move on to the simple internal routing protocols (RIP, then EIGRP).

    You;re already way ahead of a lot of people I've worked with who were (ahem) 'network engineers) :biggrin

    One more thing you'll need to remember is that, if you use a serial link between the two routers (in the real world you probably would) you'll need to clock them - take a look at some Cisco documentation for an explanation of what that means and how to implement it - though again, PT gives you a 'cheat' method of doing it if you get stuck...
     
    Certifications: A few
    WIP: None - f*** 'em
  6. zr79

    zr79 Byte Poster

    199
    1
    17
    Ok i just worked out how to do static routes, for the above 2 router scenario i just added,

    Router0
    ip route 192.168.2.0 255.255.255.0 192.168.1.2

    Router1
    ip route 192.168.0.0 255.255.255.0 192.168.1.1

    So the next hop is just the adjacent routers ip on the side you are travelling to and the network ip is the default network ip on the other side of this adjacent router, seems pretty easy.

    BUT, what if i add another router Router2 coming off Router0, now using this above static route method i can static route Router2 by adding.

    Router0
    ip route 192.168.4.0 255.255.255.0 192.168.3.2

    Router2
    ip route 192.168.0.0 255.255.255.0 192.168.3.1

    But Router1 has 2 hops to the 192.168.5.0 network and it seems all i need to do is route to the 192.168.5.0 network from Router1 and i don't have to add a static route to the 192.168.4.0 network from Router1 in other words all i need to be able to Route Router1 past Router2 onto 192.168.5.0 is,

    Router1
    ip route 192.168.4.0 255.255.255.0 192.168.1.1

    Router2
    ip route 192.168.1.0 255.255.255.0 192.168.3.1

    but



    Router1
    ip route 192.168.5.0 255.255.255.0 192.168.4.2

    Router2
    ip route 192.168.0.0 255.255.255.0 192.168.4.1

    [​IMG]
     
    Certifications: A+
  7. Simon-MCT

    Simon-MCT Bit Poster

    18
    3
    27
    As Zebulebu said above, you're getting the hang of this. I will defer to others in this matter for Cisco specifics - when I first posted I didn't realise it was the Cisco forum! I've studied the CCNA course but it's been a while since I actually did any Cisco config.

    One thing to remember with routing is that a device only needs to know it's next hop. For example to go from Router 1 to anywhere but the 2.0 subnet, it sends data to Router 0. Router 2 is the same, anywhere but the 4.0 subnet goes to Router 0.

    Router 0 is the more complex one, since it knows about the 0.0 subnet, and two other routers. This is where routing protocols come in, since those other routers will tell Router 0 which subnets they're connected to. This will let Router 0 build a routing table, knowing which ports to send data out of to get to foreign subnets.

    One thing you haven't listed in your IP static routes above is the interface that the data is to pass out of from a router. With Routers 1 and 2 this is simply the default gateway, set in global config, if I remember correctly, since there is only one way for the data to go.

    As far as your Ping question above is concerned, Ping is two-way traffic. An ICMP ping packet is sent out from the source to the destination which is probably (if routing is working) reaching it's destination. The issue usually lies with the response packet not having a route back to the originating device. Hence you can ping directly connected devices, but not "across" the routers. Some network monitoring should tell you what's happening there.

    Finally, as for my trick question about how many subnets there should be, you got it wrong with the number (you stated 2) but right with the IP addressing (subnets 0.0, 1.0 and 2.0)! A lot of folks ignore the fact that there's a subnet (or more) connecting the routers, and I defer to the more specific IP addressing or serial link as suggested by Zebulebu.
     
    Certifications: MCSE:Sec;MCITP; MCTS; MCT; A+,Sec+
    WIP: CCNA
  8. zr79

    zr79 Byte Poster

    199
    1
    17
    thanks for that, actually i did know there were 3 subnets, each segment is basically a subnet, but as i was doing it in terms of floors i left the first subnet out.
     
    Certifications: A+
  9. Simon-MCT

    Simon-MCT Bit Poster

    18
    3
    27
    Fair enough. I was just trying to illustrate the point that there can be subnets without computers on them, just connecting routers together, that's all. :biggrin

    I need to draw up a JPG of the router config I use on my courses to see what you'd make of it. From what you've posted so far I reckon you'd come up with a suitable IP addressing scheme and be able to tell me what should be in the routing tables anyway.
     
    Certifications: MCSE:Sec;MCITP; MCTS; MCT; A+,Sec+
    WIP: CCNA
  10. zr79

    zr79 Byte Poster

    199
    1
    17

    Go for it!
     
    Certifications: A+
  11. albertc30

    albertc30 Kilobyte Poster

    423
    1
    37
    Hello zr79.

    Here is one way of doing it.
    No static routes. Instead I am using routing information protocol (RIP) v2.

    I am also using variable length subnet masking so be aware that RIP v1 does not support it (VLSM), so we must use v2 by using comand #version 2 under #router rip.

    Please see attached picture and feel free to ask any questions either pm me or here to the forums as the guys and girls do help and alot.

    I know so as I am just halfway my CCNA Module 4 and I am now on PPP and the guys here have put up with alot trust me.

    Take care.
     

    Attached Files:

    Certifications: CCNA
    WIP: 220-701 - A+

Share This Page

Loading...