Various questions for CCNA (cmds, NBMA,FR,ISDN)

Mrock4

Hey guys. I'm currently in Iraq working in a network support role..though since our initial hell stage, things have calmed down, and there isn't much of a need to touch any routers or switches around here anymore. I'm currently using the Cisco press INTRO/ICND books, and the preplogic practice exams. I am going home for two weeks to visit everyone, and I arrive on the 24Nov06..and am scheduled to take my CCNA on the 27th (which I may push back a day or so..depending on my arrival).

Anyways, my questions- I am trying to get a complete grasp of everything, since the test is creeping up, and I'm not feeling too confident yet. I know a decent range of the material, but Frame Rely, OSPF on NBMA, and ISDN are hurting me right now..especially ISDN.

What my question really comes down to- what is the best way to attack the material with it coming up so soon? I am scoring on and off around 900 on my PrepLogic practice exams, but that's not a big enough margin to make me happy. Any last minute tips? Right now, I am trying to re-watch the CBT Nuggets for the CCNA, and re-read the sections I am most scared of (configuring dialer profiles kills me..). Any further suggestions?

I appreciate any advice, and I apologize if my post is somewhat cloudy..my mind is in two thousand different places at this moment. I am going to hit the books for a little while..

Mike

Edit: Upon beginning studying, I thought of a couple of questions I actually could lay out clearly.

NBMA- One of my CBT Nuggets said "OSPF is tricky over NBMA". Are all dynamic routing protocols tricky over NBMA? It makes sense to me that they would, but, I'm kind of unsure as to why they only specified OSPF.

Subnetting- I am OK with subnetting itself, I have a pretty strong grasp on it..but the one type of question I am not good at is- "You have the 172.16.0.0 address space, divide this up into ___ subnets with ___ hosts on each subnet. What is the best way to divide them up to be equal? Or the fastest, at least.

frame relay map vs. frame relay route? My ICND book shows "frame relay map ip x.x.x.x DLCI broadcast" as an example. I was looking somewhere else (I don't quite remember where), and it showed "frame relay route...". I haven't seen the command in the book that I can remember, so I'm a little lost here. What is the difference? I know the frame relay map command maps the DLCI to the IP, but what does the route command do then?

Paul Boz

the reason NBMA makes OSPF hard is because OSPF uses broadcast hello packets to keep up to date on neighbor statuses. NBMA means no broadcast though, so OSPF is forced to send hello packets to each host. As a result, the Hello Packet timer is moved from 10 seconds to 30, and the dead route timer is moved from 40 to 120 seconds.

A good rule of thumb for NBMA is to remember that if a routing protocol uses broadcasts (which most do) you'll run into trouble. thankfully, OSPF has a provision for NBMA.

Understanding subnetting is by far the most critical thing you learn from persuing a CCNA level certification. Those questions are usually pretty tough if you don't have a firm grasp on subnetting, so study as much subnetting as it takes. I read the mindleaders.com resources, read the Lamle Sybex book, and read several web resources to really get it down.

If you get a question like "You're given a class C IP range and need 16 networks with 10 hosts per network, first check to see if the question mentions whether or not IP Subnet Zero is disabled. If the question doesn't mention it assume it's enabled. Then ask yourself how many bits you have to play with. In a class C network you've got eight, so you have to think how many bits you can use for the network portion and how many you can use for hosts. In your case, a /28 network (four network bits and four host bits) would be what you need. This is because using a /28 netmask (255.255.255.240) will provide you wiith 16 networks, and using the remaining (four) its for hosts, provides 14. You'd think that it would be 16 and 16, but you have to remember that there's a gateway and broadcast address for each subnet, so you have to subtract two from your final total. Note that if IP subnet Zero is not enabled you also have to subtract two from the network number.

Here's how I'd work out the above problem logically.

You're given a class C IP range and need 16 networks with 10 hosts per network. How would you subnet this network to provide the right number of networks and hosts?"

1.) Q. Is IP Subnet Zero enabled?
A. The question doesn't mention it, and it's enabled by default, so yes.

2.) Q. How many bits do I have to play with?
A. It's a class C network (192x.x.x) so you have eight bits to play with.

3.) Q. How many bits do I need to equal the number of networks?
A. a /28 network (mask of 255.255.255.240) will work. You can tell by subtracting 240 from 256, which equals 16, giving you just enough networks. (If IP Subnet Zero was disabled you'd have to subtract two, leaving you with 14 networks).

4.) Do I have enough bits left for the correct number of hosts?
A. Yes, you have four bits left, but only need 10 hosts per network. 2^4 = 16, then subtract two (because each subnet has a gateway and broadcast address) and you wind up with 14 hosts per network, which is well within range of your requirements.

If you have any other subnetting questions or need me to explain this further let me know This method works for all classes. You should spend some time making sure that you know what the CIDR notations are so you can do this stuff in your head. By this, I mean know that a /26 netmask is 255.255.255.192, etc.

Regarding the frame relay question, I'm pretty sure that using "frame relay route" is only used when configuring the router to act as a frame relay switch.

About "frame relay route"
http://www.ciscopress.com/articles/article.asp?p=170741&seqNum=7&rl=1

about "frame relay map"
http://www.ciscopress.com/articles/article.asp?p=170741&seqNum=6&rl=1

loboernesto

frame relay map vs. frame relay route?

I think ( not sure..I also have to review WANs before i take the exam), that you use frame relay route when you configure a router to act as a frame-relay switch..if I'm not wrong (this is coming out of my head right know..) this will route packets between DLCIs:

frame-relay route DLCI interface DLCI

example:
router1(config)#interface serial1
router1(config-if)#frame-relay route 102 Interface serial2 201

this will route traffic entering interface s1 with DLCI 102 out interface s2 with DLCI 201.

if you have a chance to configure a frame-relay switch there's an example of the config in thebryantadvantage.com which I used and works well. Configuring a frame-relay switch really helps you understand frame-relay, it did to me at least.
cheers

Mrock4

Paul- The NBMA/OSPF explanation helped..I didn't know anything about the Hello/dead timers. Good timing, considering tonight I played with the hello/dead timers on my 3 2600's before coming to work.

The subnetting explanation you gave was pretty much all things I felt I knew already, so I guess my best plan of attack is to just write it all down a little more. I tend to think about it for a few seconds, and move on, as opposed to writing it down. I'll do a few practice problems and review what you said again. Thanks!

lobo- I see what you mean. That makes complete sense. I should have seen the difference but really didn't look at the FR route command long enough. Thank you also..


I am feeling somewhat more confident after my little session today. I have a 2620 at my disposal, so I went and borrowed two 2600's from a buddy of mine to work on. I sat for a few hours configuring them to route RIP, EIGRP, then OSPF, and then screwing up the configurations to see what effect it would have on other routers. That is how I ended up playing with the hello/dead timers. I also went through a CCNP- CIT CBT Nugget on troubleshooting OSPF, and created problems for myself, and explored through the various configs to find a solution. If I could get that much hands on each day until the test I would be completely confident. I think FR is something I can memorize fairly well, but ISDN is going to kill me.

Question: I don't think it is on the CCNA, but I want to learn what I can. In the interface command, I set ip ospf authentication ..on one of the OSPF interfaces, to see what the debug command would bring for the mismatch of one interface requiring authentication, and the other not..however there was no problem, the neighbor relation seemed to stay at full, and I had complete connectivity. What gives? I had also used the "area 0 authentication message-digest" command in router config..and this *did* cause a drop in the neighbor relationship. Why did that cause a problem, but the interface authentication didn't? I may not have gone through all of the necessary steps.

Anyways, I've got to get back to work, SNMPc is showing too much red! I apologize for the ridiculous amount of questions, I know it gets old.

-Mike

Paul Boz

Oh no, it's fine man, the questions are totally fine

Kaminsky

Mrock4 wrote:

I'll do a few practice problems and review what you said again. Thanks!

-Mike

With these questions move straight to the binary.. it's the quickest way. How many bits do you need for x subnets. How many for hosts. Then fit it around the IP Class that you have.

If you move straight to binary (or nnnnn hhhhhhh) the questions works itself out.

Look here for subnetting practice - http://www.subnettingquestions.com/default_uk.asp

Spend several hours on there going over and over and you will be swatting the subnet questions like flies.

gluck with the exam btw.[/url]

Danman32

Try to memorize the decimal numbers of common subnet mask octets.
128 - 1 bit
192 - 2 bits
224 - 3 bits
240 - 4 bits
248 - 5 bits
252 - 6 bits
254 - 7 bits
255 - 8 bits

Mrock4

I think I am good to go now. I was just looking at the "find how many subnets you need for x amount of hosts" question wrong. I took a practice test with somewhere around 5 or 6 questions of that sort alone, and nailed each one pretty quickly. The solution? I just wrote it down instead. On the more basic ones such as "You need to support x amount of users, what subnet mask will you need?", I can generally say off the top of my head. I guess I need to make it a habit to stop underestimating those and actually verify my answers on paper before I make them final.

I appreciate all of the help I've received so far. I'm definently more confident on that aspect of the test at least, and very impressed at how many replies I've gotten on this. Very awesome! I'm sure you'll hear from me again shortly as I'll be taking frame relay head on shortly. Thanks again guys.

-Mike