A couple of questions

the_dude7the_dude7 Posts: 31Registered Members ■■□□□□□□□□
I'd just like a few things explained: I'm doing some things that according to the book I should not, but they seem to be okay

1) From what I understood, you had to use crossover Ethernet cables to connect different switches together. However, I'm using regular straight-throughs for that in my home lab with my equipment and it still seems to work okay; I can ping fine between them and everything, and work with STP, VTP, VLANs, EtherChannel, etc. And I have connectivity all the way from my host laptops through the switches and routers to the other end of the network. How is this possible?

2) We know what the three private address ranges are. However, for my home lab, I'm just assigning to some interfaces other random IP numbers outside that range for certain occasions (like 40.40.40.1), and it still works fine. Do the private address ranges only apply to networks that have a connection to the Internet and thus in the context of NAT? As in if you want a private address translated, it has to be within those ranges like 192.168....

Comments

  • clarsonclarson Posts: 875Registered Members ■■■■□□□□□□
    1) auto mdix

    2) you probably been allowed use of a public ip address by your isp. then you nat your private addresses to the internet. you could still use another public ip address, but your isp isn't going to allow it on it's network. so it won't ever reach an internet address. if your just connecting your lab equipment together and it isn't going to the internet, it doesn't matter what addresses you use
  • the_dude7the_dude7 Posts: 31Registered Members ■■□□□□□□□□
    Thank you. I wish someone had told me about Auto-MDI-X before I went and bought crossover cables, since I had tons of regular straight-through twisted pairs already. I guess they put that in the curriculum because it's how things used to be done, and in case you work with older equipment that doesn't have auto negotiating on there? Or if that's disabled.
  • clarsonclarson Posts: 875Registered Members ■■■■□□□□□□
    well you could have tested a straight through cable before hitting the buy now button.

    while newer switches almost always support auto mdix, there are a lot of devices that don't. think the internet of things. and manufacturers desire to make things cheaper so they make more money.
  • Jon_CiscoJon_Cisco Posts: 1,556Registered Members ■■■■■□□□□□
    It's all part of the learning process. You will probably have no problem understanding auto mdix now since you had a real life experience to relate to. For troubleshooting and for test purposes you will want to know why a straight thru and crossover cable works. It has to do with the send and receive pins. Some devices are smart enough to adjust this based on the cable type.

    For the IP question your home network devices don't know or care what network you use. If you assign someone else public IP on your internal network nobody will know or care. However your external IP is provided by your service provider. That one must be correct and that it the IP that you would use when performing NAT.

    You could run into some strange routing issues if you use a public IP on your internal network and then try to connect to the same public ip on an external network. Your router will not sent the traffic to your default gateway because it will believe you are already on that network. In the end have fun playing and testing because that is the best way to learn.
  • awitt11awitt11 Posts: 50Registered Members ■□□□□□□□□□
    What devices and software versions are you running? Something else to keep in mind. While auto-MDX is great for switches, you may still need the crossovers as you experiment with routers.
  • the_dude7the_dude7 Posts: 31Registered Members ■■□□□□□□□□
    ^Indeed that is certainly true. For routers I still need them, so I didn't waste cash. Although I'm ordering more serial connectors for them as well.

    They were both stupid questions and I feel ashamed for asking them. Had I but researched the issue a bit beforehand I would've found out icon_sad.gif

    ----

    However here's another question since I don't want to flood this forum with new threads:

    When setting up authentication with RADIUS or TACACS+, for console and VTY access for example... I know that first you have to type

    "aaa new-model" and then you do "radius server SecureLogin" for ex., and then have to input an address:

    "address ipv4 . . . ." and "key MyRadiusPassword" for example, and then add to an AAA group of any name

    "aaa group server radius MyRadiusGroup"
    "server name SecureLogin"

    and finally

    "aaa authentication login default group MyRadiusGroup local"

    This is what the Lammle book says, but what are you supposed to put in for the IP address there? Is the actual RADIUS authentication server a separate computer somewhere on the network that has that program? Or by server does he mean the router you're on that you're essentially configuring to act like a server (in this regard/for this purpose), and it already has these authentication capabilities built in? This isn't explained at all and it's making me angry.
Sign In or Register to comment.