Hi xvier1, you do have an idea about networking..cisco related and all ? A good and reliable internet for video sessions ?

xvier1:

no problem i will really appreciate that. just let me know wen the session is scheduled. my mail is xakpan@yahoo.com

Hello Guys, just decide to share an Enterprise network project which I work on last year. In also includes VOIP loads of other stuff, redistribution btw eigrp, ospf and rip. ACls, portsecuirty, NAT, FrameRelay, Router on a stick. etc

Your questions are welcomed.

snipesdam:
Hello Guys, just decide to share an Enterprise network project which I work on last year. In also includes VOIP loads of other stuff, redistribution btw eigrp, ospf and rip. ACls, portsecuirty, NAT, FrameRelay, Router on a stick. etc

Your questions are welcomed.

Your pix is kinda tiny and I can't make out anything on it to ask or make a contribution on it. Can you upload a better quality image please?

Just took a second look at it, and I must say that for an enterprise network, it is poorly designed. Just looking at the network segment on the left is giving me a headache. And you mentioned doing a router on a stick, with all these devices you have in place here, why would you want to do an ROAS? That is for a small sized network with fewer devices. On yours, you have enough to achieve redundancy and no need for an ROAS. You actually tried to go for redundancy on the left segment but you went overboard with it. Same goes for the network at the bottom.

henryskywalker2003:
Just took a second look at it, and I must say that for an enterprise network, it is poorly designed. Just looking at the network segment on the left is giving me a headache. And you mentioned doing a router on a stick, with all these devices you have in place here, why would you want to do an ROAS? That is for a small sized network with fewer devices. On yours, you have enough to achieve redundancy and no need for an ROAS. You actually tried to go for redundancy on the left segment but you went overboard with it. Same goes for the network at the bottom.

Perhaps you haven't experimented on redundancy on large scale. Am not disappointed that you said it poorly designed only if saw the running config. About the ROAS its a design which i choose inorder to implement VTP and ROAS isn't just done for small networks. Dunno why the image is blurry tried making better.

Also below you can find an experiment on split-horizon which i designed last xmas. Its frameRelay multipoint and branch can only be neighbors with the core via EIGRP routing protocol. But they are able to get update of the other branches subnets but NOT Neighbors.

Hi snipesdam,

Thanks for sharing your topology. How are you able to emulate IP phones on PacketTracer ? Or, are those analog phones ? I'm taking the Voice and I know for sure simulation doesn't come close to labbing for voice. Good job on the topo though.
You connect everything like Barb wire haha

snipesdam:
Hello Guys, just decide to share an Enterprise network project which I work on last year. In also includes VOIP loads of other stuff, redistribution btw eigrp, ospf and rip. ACls, portsecuirty, NAT, FrameRelay, Router on a stick. etc

Your questions are welcomed.

You are somewhat correct on the idea of router-on-a-stick but for the sake of labbing I don't think there is a limit on to how many routers you want to configure or else if this topo is designed for a real live network site. And by the redudancy, You never enough to achieve redundancy o. You never know what links or what ports may go at whatever time.

henryskywalker2003:
Just took a second look at it, and I must say that for an enterprise network, it is poorly designed. Just looking at the network segment on the left is giving me a headache. And you mentioned doing a router on a stick, with all these devices you have in place here, why would you want to do an ROAS? That is for a small sized network with fewer devices. On yours, you have enough to achieve redundancy and no need for an ROAS. You actually tried to go for redundancy on the left segment but you went overboard with it. Same goes for the network at the bottom.

DanIndia:
You are somewhat correct on the idea of router-on-a-stick but for the sake of labbing I don't think there is a limit on to how many routers you want to configure or else if this topo is designed for a real live network site. And by the redudancy, You never enough to achieve redundancy o. You never know what links or what ports may go at whatever time.

I see you understand redundancy better . In real life I've had an experience where a port went bad and you wouldn't know immediately where the issue was coming from. Also yea the redundancy may look like barb wire it's pertinent to design for worse case scenarios for real live and the idea is a MESH like redundancy and you have to choose the Root Bridge carefully to make sure STP or RSTP runs effectively.

About VOIP . Yes I successfully implemented it on Packet Tracer and it worked very well they phones rang. Assigned phone numbers, ip, telephony service, but in real live you will have to TFTP to retrieve IP phones config etc. Even my tutor back then when I was learning CCNA was impress, he wasn't able to do it. But note that you can't do VOIP routing.

So when designing yours think like a network architect . Never design for the present accommodate it for expansion. Think out of the box and seek for improvement from peoples ideas too. You can as well tell me what I missed and make my topology better.

snipesdam:


Perhaps you haven't experimented on redundancy on large scale. Am not disappointed that you said it poorly designed only if saw the running config. About the ROAS its a design which i choose inorder to implement VTP and ROAS isn't just done for small networks. Dunno why the image is blurry tried making better.

Also below you can find an experiment on split-horizon which i designed last xmas. Its frameRelay multipoint and branch can only be neighbors with the core via EIGRP routing protocol. But they are able to get update of the other branches subnets but NOT Neighbors.

Hi Snipesdam

Lol. I haven't "experimented" with redundancy on a large scale but I have actually implemented that. Now let me give you reasons why I say it is poorly designed based on what I am seeing on your diagram.

1. No one uses ROAS on an enterprise network if you want to implement redundany. If you can think back on the concept of ROAS, the vlans get routed using a single device which in your case is the router. Now the router isn't as fast as a switch is because it is software based unlike the switch that uses the ASIC which is a hardware implementation. Again, the way the ROAS works means that all the traffic comes to the router and exits it on a single port. Now imagine having a large amount of traffic coming and leaving the router on a 100mbps or 1gbps link. So, you will agree with me that ROAS isn't the way to go in this case.

2. Strictly speaking, I think you designed the network on the left to look good rather than to be functional, same thing for the bottom one. I mean, take a look at the way you arranged your switches. You didn't follow the basic hierachy of design which means that you should have an Access, Distribution and Core Layers, or at least have a Collapsed Core where you merge your Distribution and Core Layers. When STP kicks in, which it will or else you will have loops, All your redundant links will go down and you will effectively have only single working links at any given time, because the redundant ones will be blocked and hence not being used. So, in essence, you have a couple of links siting there doing nothing.

3. STP issues. Still on the network on the left, now, I don't know the config on those devices, but let's assume that the switch closest to the router gets elected to be the root bridge. Now, unless you specifically tweaked it, you know that what follows can be very random. Every switch finds the best way to get to the root bridge and still does that based on the same way that the root bridge is elected which means the oldest device/ lowest priority/lowest port number. So, imagine if the leftmost switch has a lower priority/mac/port number than that of the two switches connected to it and they in turn are lower than the other two, that means that the two switches closest to the root bridge will end up going all the way back to the leftmost switch in other to reach a root bridge that is directly connected to it. So, you need to re-arrange ur switches and optimise them for STP, which brings me to the next reason

4. STP presence. Looking at that network, it is obvious that you are using STP on the network. If you want a network with redundancy and high availability and very fast failovers, then you need to get rid of STP totally. This is 2015 for crying out loud, STP takes at best 15-30secs to failover. That is not good at all for any enterprise network. Even Cisco recommends that you get rid of STP on your network unless you ABSOLUTELY need it. So, designing an enterprise network with STP in mind and ending up designing it poorly isn't good at all.

5. Redundancy for the sake of redundancy is bad. You don't just start throwing redundant links just for the sake of it. Looking at your links, you even connected Access switches together which is also a terrible design. Access switches should ALWAYS be separate from each other unless they are feeding the same Vlans/subnet.

6. IP phones were eating up network ports. Why give your ip phones and PCs different ports when you can easily use one port on ur switch for an ip phone and a pc. The ip phones have an in built switch so, you can plug in your pc into the back of the ip phone and while the phone gets pluged into the switch. In the basic form, you can configure the native vlan of the switch to be that of the pc to still have your vlans working and if your switch supports it, you can configure a voice and access vlan on a single port.

I will like to see your configs on your network and also you ip address scheme.

DanIndia:
You are somewhat correct on the idea of router-on-a-stick but for the sake of labbing I don't think there is a limit on to how many routers you want to configure or else if this topo is designed for a real live network site. And by the redudancy, You never enough to achieve redundancy o. You never know what links or what ports may go at whatever time.

@DanIndia, there is a limit when it comes to redundancy on a network. You don't go out to complicate your network just so that you can have redundancy. There should be a limit that when you go beyond, it starts impacting on your network's functionality and perfomance. Imagine doing a full mesh of 6 switches all in the name of redundancy. That is chaos for any network admin. Take a look at the network I posted. The Access switches have redundant links to each of the Distribution switches. The four links cannot suddenly go down at the same time. There is no ROAS, rather the two Distribution switches act as a gateway to the access switches at the same time. The router, then acts as a gateway for the switches and up there, it is purely L3 no more L2. The switches are connected to the router using L3 links and OSPF routes traffic between the 3. The router advertises a default gateway to the two switches (NOT BY REDISTRIBUTION) so the link from the router to any of the switches goes down, the other switch still gets the advertisement and forwards it to the switch that lost it's connection to the router.

@Henry, you really think there is a limit as to redundancy ? No, I don't think so. Not at all... and I don't think his topo is messed up at all. Have you seen CCNP lab topos ? There is no breathing space at all. I mean everything is connected to everything... and yes, I've seen live setups with even more redundancy. 6-7 years ago, you could say there is a limit but not these present days when a delay of secs is not permissible. You are kind of comparing your own topo with his, remember, this is a learning platform where people can mess things up, fix it again.. there is never a stage where you stop, you always, always go beyond. I'll try and get some CCNP topologies and post it here - then you'll understand why I said his topo isn't messed up at all. Why do you think Cisco introduced Eterchannel ? You haven't seen just two routers connected with like 5-6 links ? All for the sake of redundancy ? If you go VoIP, you'll get to see how many connections people use just to connect to a CO. Imagine, having one line through to a PBX and it's for a company, while that line is being used, all other lines are gonna get dropped and this may seem to be a little bit different from the current scenario its another form of redundancy.

henryskywalker2003:


@DanIndia, there is a limit when it comes to redundancy on a network. You don't go out to complicate your network just so that you can have redundancy. There should be a limit that when you go beyond, it starts impacting on your network's functionality and perfomance. Imagine doing a full mesh of 6 switches all in the name of redundancy. That is chaos for any network admin. Take a look at the network I posted. The Access switches have redundant links to each of the Distribution switches. The four links cannot suddenly go down at the same time. There is no ROAS, rather the two Distribution switches act as a gateway to the access switches at the same time. The router, then acts as a gateway for the switches and up there, it is purely L3 no more L2. The switches are connected to the router using L3 links and OSPF routes traffic between the 3. The router advertises a default gateway to the two switches (NOT BY REDISTRIBUTION) so the link from the router to any of the switches goes down, the other switch still gets the advertisement and forwards it to the switch that lost it's connection to the router.

I'm running out for a bit but I thought I'd comment on your point No. 6.
I'd rather lay two cables to connect my PC and IP phones rather than connecting a PC to the switchport of an IP phone. If you've ever troubleshooted a complex network, you'll understand why I said so

henryskywalker2003:


Hi Snipesdam

Lol. I haven't "experimented" with redundancy on a large scale but I have actually implemented that. Now let me give you reasons why I say it is poorly designed based on what I am seeing on your diagram.

1. No one uses ROAS on an enterprise network if you want to implement redundany. If you can think back on the concept of ROAS, the vlans get routed using a single device which in your case is the router. Now the router isn't as fast as a switch is because it is software based unlike the switch that uses the ASIC which is a hardware implementation. Again, the way the ROAS works means that all the traffic comes to the router and exits it on a single port. Now imagine having a large amount of traffic coming and leaving the router on a 100mbps or 1gbps link. So, you will agree with me that ROAS isn't the way to go in this case.

2. Strictly speaking, I think you designed the network on the left to look good rather than to be functional, same thing for the bottom one. I mean, take a look at the way you arranged your switches. You didn't follow the basic hierachy of design which means that you should have an Access, Distribution and Core Layers, or at least have a Collapsed Core where you merge your Distribution and Core Layers. When STP kicks in, which it will or else you will have loops, All your redundant links will go down and you will effectively have only single working links at any given time, because the redundant ones will be blocked and hence not being used. So, in essence, you have a couple of links siting there doing nothing.

3. STP issues. Still on the network on the left, now, I don't know the config on those devices, but let's assume that the switch closest to the router gets elected to be the root bridge. Now, unless you specifically tweaked it, you know that what follows can be very random. Every switch finds the best way to get to the root bridge and still does that based on the same way that the root bridge is elected which means the oldest device/ lowest priority/lowest port number. So, imagine if the leftmost switch has a lower priority/mac/port number than that of the two switches connected to it and they in turn are lower than the other two, that means that the two switches closest to the root bridge will end up going all the way back to the leftmost switch in other to reach a root bridge that is directly connected to it. So, you need to re-arrange ur switches and optimise them for STP, which brings me to the next reason

4. STP presence. Looking at that network, it is obvious that you are using STP on the network. If you want a network with redundancy and high availability and very fast failovers, then you need to get rid of STP totally. This is 2015 for crying out loud, STP takes at best 15-30secs to failover. That is not good at all for any enterprise network. Even Cisco recommends that you get rid of STP on your network unless you ABSOLUTELY need it. So, designing an enterprise network with STP in mind and ending up designing it poorly isn't good at all.

5. Redundancy for the sake of redundancy is bad. You don't just start throwing redundant links just for the sake of it. Looking at your links, you even connected Access switches together which is also a terrible design. Access switches should ALWAYS be separate from each other unless they are feeding the same Vlans/subnet.

6. IP phones were eating up network ports. Why give your ip phones and PCs different ports when you can easily use one port on ur switch for an ip phone and a pc. The ip phones have an in built switch so, you can plug in your pc into the back of the ip phone and while the phone gets pluged into the switch. In the basic form, you can configure the native vlan of the switch to be that of the pc to still have your vlans working and if your switch supports it, you can configure a voice and access vlan on a single port.

I will like to see your configs on your network and also you ip address scheme.

DanIndia:
@Henry, you really think there is a limit as to redundancy ? No, I don't think so. Not at all... and I don't think his topo is messed up at all. Have you seen CCNP lab topos ? There is no breathing space at all. I mean everything is connected to everything... and yes, I've seen live setups with even more redundancy. 6-7 years ago, you could say there is a limit but not these present days when a delay of secs is not permissible. You are kind of comparing your own topo with his, remember, this is a learning platform where people can mess things up, fix it again.. there is never a stage where you stop, you always, always go beyond. I'll try and get some CCNP topologies and post it here - then you'll understand why I said his topo isn't messed up at all. Why do you think Cisco introduced Eterchannel ? You haven't seen just two routers connected with like 5-6 links ? All for the sake of redundancy ? If you go VoIP, you'll get to see how many connections people use just to connect to a CO. Imagine, having one line through to a PBX and it's for a company, while that line is being used, all other lines are gonna get dropped and this may seem to be a little bit different from the current scenario its another form of redundancy.

@DanIndia, you are talking about labs, I'm talking about live topologies. You can go as crazy as you want in a lab topology, but when you try that in live topologies, that's when you run into issues. I myself have also seen live equipment with more redundancy than usual, but these are very large enterprise networks. I'm talking about ISPs here and really large organisations. As for the routers connected by 5-6 links, I can bet you that the network supports thousands of devices at the Access layer. Besides, you didn't even say anything about the network lacking a logical, hierachical structure. That alone shows a very poor design. Imagine trying to troubleshoot such a network or when you need to scale out the network. Where do you even start to add new switches when you want to scale out? As for a network where everything is connected to each other, that is one hell of a terrible network, you can quote me on that anytime, anyday. Everything is not supposed to be connected to everything. Instead, every device on a layer must be directly connected to every device on the layer above and below it. Hence, every access layer device gets connected to every distribution layer device but not the core or edge. The distribution layer gets connected to both the core and access but not the edge and so on. If you connect everything to everything how do you configure your security permissions and restrictions, QOS, etc. I actually studied CCDA as well as CCNA so, not just saying this for the sake of saying it.

Read these articles to help you understand what I am trying to say here.

http://www.ciscopress.com/articles/article.asp?p=2189637&seqNum=4

http://www.nojitter.com/post/240151667/network-redundancy-or-resilience

Bro mr danindia pls am just goin tru nairaland I saw ur thread pls dis wat I been lookin for de past 2yrs now pls I want to do my IT program in india I need all available infor abt pls I will be grad if u can help me by send all de requirements nd I will like to have ur number pls tanx my email is icezychris@yahoo.com. Sorry for interuptin ur conversations

Please stop spamming all threads, please.!

icezychris:
Bro mr danindia pls am just goin tru nairaland I saw ur thread pls dis wat I been lookin for de past 2yrs now pls I want to do my IT program in india I need all available infor abt pls I will be grad if u can help me by send all de requirements nd I will like to have ur number pls tanx my email is icezychris@yahoo.com. Sorry for interuptin ur conversations

You still don't get my point. What I'm saying is don't ridicule him. He did his by sharing his topology. It could be for a site and could be for labbing purposes. And by the way, your idea is an opinion, not a fact. People can connect their devices however they want. I understand you're explaining about core layer and access layers, we all are familiar with those terms.. those are somewhat at beginner's level. The guy who posted the initial topo was just looking to share. He never asked to be talked down on. For all I can gather, I don't think he implemented that on a live network and when he posted, he asked for ideas on how to make it better (I believe).
And believe me, you'll find really, really messy labs. Walk into some offices, you'd hate the admin for life, but still it runs and everything works just fine. I can tell you this from what I've seen. The lab isn't perfect neither is it that bad.
And if you take a closer look at the lab, I don't see so much loopholes in it, it's pretty neat. The left side devices are switches not routers and we both know switches are dummy devices, they are not as smart as L3 devices and hence VLANS and Switchport securities can take care of any form of threats posed to that part of the topology. The HQ consists of 4 routers which are inter-connected and are not connected to any devices that are outside the main site. The only drawback I noticed was router-on-a-stick which is not mandatory because of the devices involved and that again, doesn't really matter on how a designer would want his network to be. Not everything goes but as far as this topo is concerned, everything is connected to everything. And this term does not indicate you connect a PC to the core devices. From a networking term, I hope you get what I'm saying.

henryskywalker2003:


@DanIndia, you are talking about labs, I'm talking about live topologies. You can go as crazy as you want in a lab topology, but when you try that in live topologies, that's when you run into issues. I myself have also seen live equipment with more redundancy than usual, but these are very large enterprise networks. I'm talking about ISPs here and really large organisations. As for the routers connected by 5-6 links, I can bet you that the network supports thousands of devices at the Access layer. Besides, you didn't even say anything about the network lacking a logical, hierachical structure. That alone shows a very poor design. Imagine trying to troubleshoot such a network or when you need to scale out the network. Where do you even start to add new switches when you want to scale out? As for a network where everything is connected to each other, that is one hell of a terrible network, you can quote me on that anytime, anyday. Everything is not supposed to be connected to everything. Instead, every device on a layer must be directly connected to every device on the layer above and below it. Hence, every access layer device gets connected to every distribution layer device but not the core or edge. The distribution layer gets connected to both the core and access but not the edge and so on. If you connect everything to everything how do you configure your security permissions and restrictions, QOS, etc. I actually studied CCDA as well as CCNA so, not just saying this for the sake of saying it.

Read these articles to help you understand what I am trying to say here.

http://www.ciscopress.com/articles/article.asp?p=2189637&seqNum=4

http://www.nojitter.com/post/240151667/network-redundancy-or-resilience

And keep in mind that I don't disagree with almost everything you've mentioned in all of your posts. In fact, I think you'll make one fine network designer. You seem to like everything spick and span

henryskywalker2003:


@DanIndia, you are talking about labs, I'm talking about live topologies. You can go as crazy as you want in a lab topology, but when you try that in live topologies, that's when you run into issues. I myself have also seen live equipment with more redundancy than usual, but these are very large enterprise networks. I'm talking about ISPs here and really large organisations. As for the routers connected by 5-6 links, I can bet you that the network supports thousands of devices at the Access layer. Besides, you didn't even say anything about the network lacking a logical, hierachical structure. That alone shows a very poor design. Imagine trying to troubleshoot such a network or when you need to scale out the network. Where do you even start to add new switches when you want to scale out? As for a network where everything is connected to each other, that is one hell of a terrible network, you can quote me on that anytime, anyday. Everything is not supposed to be connected to everything. Instead, every device on a layer must be directly connected to every device on the layer above and below it. Hence, every access layer device gets connected to every distribution layer device but not the core or edge. The distribution layer gets connected to both the core and access but not the edge and so on. If you connect everything to everything how do you configure your security permissions and restrictions, QOS, etc. I actually studied CCDA as well as CCNA so, not just saying this for the sake of saying it.

Read these articles to help you understand what I am trying to say here.

http://www.ciscopress.com/articles/article.asp?p=2189637&seqNum=4

http://www.nojitter.com/post/240151667/network-redundancy-or-resilience

DanIndia:
You still don't get my point. What I'm saying is don't ridicule him. He did his by sharing his topology. It could be for a site and could be for labbing purposes. And by the way, your idea is an opinion, not a fact. People can connect their devices however they want. I understand you're explaining about core layer and access layers, we all are familiar with those terms.. those are somewhat at beginner's level. The guy who posted the initial topo was just looking to share. He never asked to be talked down on. For all I can gather, I don't think he implemented that on a live network and when he posted, he asked for ideas on how to make it better (I believe).
And believe me, you'll find really, really messy labs. Walk into some offices, you'd hate the admin for life, but still it runs and everything works just fine. I can tell you this from what I've seen. The lab isn't perfect neither is it that bad.
And if you take a closer look at the lab, I don't see so much loopholes in it, it's pretty neat. The left side devices are switches not routers and we both know switches are dummy devices, they are not as smart as L3 devices and hence VLANS and Switchport securities can take care of any form of threats posed to that part of the topology. The HQ consists of 4 routers which are inter-connected and are not connected to any devices that are outside the main site. The only drawback I noticed was router-on-a-stick which is not mandatory because of the devices involved and that again, doesn't really matter on how a designer would want his network to be. Not everything goes but as far as this topo is concerned, everything is connected to everything. And this term does not indicate you connect a PC to the core devices. From a networking term, I hope you get what I'm saying.

@DanIndia, WOW. For real I never ridiculed him in anyway nor talked down to him. He posted that and said it was an Enterprise network project (not lab) that he worked on last year. He concluded that with saying that questions are welcome. I then made an observation which any network admin would see at first sight about the design of the network. If you notice, I didn't comment on the second topology yet because he said it was an experiment and I felt I should address the project first since it is of higher priority than the experiment. No one would like to deploy a network today only to be called back in few months time due to performance issues. And that network, if it was really a large scale network is just an issue away from a catastrophic disater. My aim was to point out the flaws there so that he can correct them and learn from it in future deployments. True, people can connect their devices how they want, but that doesn't mean they will all be the same performance-wise. Just like you could create a patch cable with any combination you like but don't expect your cable to achieve the same 100 meters expected of a cable that followed the standard. The same applies to network design. A network poorly desgined from the beginning will only lead to future problems that you aren't supposed to have.

On your last but one comment, you made a very wrong statement and I quote "The left side devices are switches not routers and we both know switches are dummy devices, they are not as smart as L3 devices and hence VLANS and Switchport securities can take care of any form of threats posed to that part of the topology.". I can't believe you just said that bro. Wow, where do I even begin to talk about this one? I hope you know that there is a huge difference between a hub, bridge and a switch? I hope you know that there are multilayer and L3 switches? Capable of almost everything a router can do? Wow. I think we need to really do a hangout very soon. Honestly, if you are studying for CCNA you need to really brush up your knowledge on Switches and Switching cos they are quite more than you think they are. The entire network depends on the switches and if your switch design is bad, your network is bad. The router does very little in a network compared to switches. Anyways, I remember you suggesting watching videos on youtube for CCNA, I highly recommend Jeremy's series from CBTNuggets. It will blow your mind.

P.S. Note that I am not talking down or ridiculing you or anything. I would really hate to see a friend pay hard earned money to write an exam only to fail it because of the wrong ideas or knowledge. I have failed an interview for a job at a big firm because of poor network design and since then, I have always strived to be better than I was then, certification or no certification.

@Henry,
I am not the kind of person that would begin to list out experiences or argue about ideas. I don't know how clarify your recent post but from what you've said... I believe you've just recently got trained in CCNA or plan to write it.
OK, let's begin with your statement - differences between hub, bridge and switches. Just so you know, hubs do not exist anymore. My first learning experience was on a HUB and this was way, way back. People do not use them anymore unless you want to blow out your network. Switch is really the same thing as a Bridge. Switches run on Bridging technology, except for the idea that the packet switching technology was introduced and read my statement clearly - Switches are generally L2 devices, their pdus are fragments... only a few switches are on L3 (i.e Multi-layer switches). And remember, from my comment I said Switches are not as smart as L3 devices. When a client calls you about a problem on L2, you directly know it's either a switch or a bridge, hub wouldn't been here but it operates on bits thus it is classified L1. If someone issues in their L3 and Switches are involved, they will clearly tell you it's a Multilayer switch. This is what is not included in most cheatsheets.

Really ? It's good to get a recommendation from you though but I've followed tons of Networking Experts - Jeremy, Brian, Kevin, INE folks, name them. I've seen them all. Oh, perhaps, you must have thought because I hadn't written my CCNA yet, you're talking to some newbie.. Nah, I've actually gone through the whole thing quite a number of times, I only didn't write it because I am matching with college timings.

Honestly, if you are studying for CCNA you need to really brush up your knowledge on Switches and Switching cos they are quite more than you think they are. The entire network depends on the switches and if your switch design is bad, your network is bad. The router does very little in a network compared to switches.
1. A Router does not do very little in a network. In fact, you can make-do with a Router only in a network (small one), and forget about switches. Haven't you heard of modules ? Perhaps, you've learnt about ISR and Modular Routers, you can integrate switchports to a Router and throw your switch somewhere. Don't tell anyone you said so. A Router does 80% of your network load. A switch only switches to the right address in your LAN which is based on mac-addresses on it's table.
2. Point 1 bases on this. Your entire network doesn't depend on Switches.

I would really hate to see a friend pay hard earned money to write an exam only to fail it because of the wrong ideas or knowledge. I have failed an interview for a job at a big firm because of poor network design and since then, I have always strived to be better than I was then, certification or no certification.
What makes you think I am going to fail the exam because of what you classify wrong ideas?! I don't think this attitude is right. Just because you failed an Interview does not mean there are not tons of people who are brighter than you are, and are fit for the said position. I wouldn't advise you to keep on with such attitude. It's a know-it-all attitude. I've been in Networking forums in which Internetworking Experts (CCIEs) are amongst and I don't recall an experience whereby an Expert talks down at people sharing ideas. For a learning platform as this, your tone of comment seems a little bit harsh, don't you think ?

And people fail exams, interviews. It's nothing new. I watched a Hangout Session with Jeremy Coiara once and someone asked, if he failed any exam and how many. His answer was "a lot". So if Jeremy could have failed exams, I don't think you failing an Interview would be anything new.

henryskywalker2003:


@DanIndia, WOW. For real I never ridiculed him in anyway nor talked down to him. He posted that and said it was an Enterprise network project (not lab) that he worked on last year. He concluded that with saying that questions are welcome. I then made an observation which any network admin would see at first sight about the design of the network. If you notice, I didn't comment on the second topology yet because he said it was an experiment and I felt I should address the project first since it is of higher priority than the experiment. No one would like to deploy a network today only to be called back in few months time due to performance issues. And that network, if it was really a large scale network is just an issue away from a catastrophic disater. My aim was to point out the flaws there so that he can correct them and learn from it in future deployments. True, people can connect their devices how they want, but that doesn't mean they will all be the same performance-wise. Just like you could create a patch cable with any combination you like but don't expect your cable to achieve the same 100 meters expected of a cable that followed the standard. The same applies to network design. A network poorly desgined from the beginning will only lead to future problems that you aren't supposed to have.

On your last but one comment, you made a very wrong statement and I quote "The left side devices are switches not routers and we both know switches are dummy devices, they are not as smart as L3 devices and hence VLANS and Switchport securities can take care of any form of threats posed to that part of the topology.". I can't believe you just said that bro. Wow, where do I even begin to talk about this one? I hope you know that there is a huge difference between a hub, bridge and a switch? I hope you know that there are multilayer and L3 switches? Capable of almost everything a router can do? Wow. I think we need to really do a hangout very soon. Honestly, if you are studying for CCNA you need to really brush up your knowledge on Switches and Switching cos they are quite more than you think they are. The entire network depends on the switches and if your switch design is bad, your network is bad. The router does very little in a network compared to switches. Anyways, I remember you suggesting watching videos on youtube for CCNA, I highly recommend Jeremy's series from CBTNuggets. It will blow your mind.

P.S. Note that I am not talking down or ridiculing you or anything. I would really hate to see a friend pay hard earned money to write an exam only to fail it because of the wrong ideas or knowledge. I have failed an interview for a job at a big firm because of poor network design and since then, I have always strived to be better than I was then, certification or no certification.

Ok. If u say so.

Alright.
I apologize if I've offended you in anyways. I don't mean to. This thread was created with a notion for people to study together. Correct people with love (church mind :p) and learn from helping people. You sure know something, and I believe others could gain from what you know. If you're up for a Hangout Session one of these days, let me know when. Maybe we could arrange a session. What you say ?

henryskywalker2003:
Ok. If u say so.

Well well well...........

Will attended to you guys this weekend. Currently busy @work.

Am interested maybe not for today buh for other days to enable me setup mine.

get me on this Onyekeluchukwuebuka@gmail.com

DanIndia:
Send me your email address

Wud appreciate it if you can also send me the link and also I wud like to join d grp!email: okunlola.israel@gmail.com

Hi Larry,

Have any basic networking Skills ? Reliable internet for video sessions ? The hangouts will resume fully sometime in the mid of next month. So you have enough time for secure a good internet access if you don't have one now.

issylarry:

Wud appreciate it if you can also send me the link and also I wud like to join d grp!email: okunlola.israel@gmail.com

Hi there,

Have any basic networking Skills ? Reliable internet for video sessions ? The hangouts will resume fully sometime in the mid of next month. So you have enough time for secure a good internet access if you don't have one now.

NOBLEDANDY:
Am interested maybe not for today buh for other days to enable me setup mine.

get me on this Onyekeluchukwuebuka@gmail.com

DanIndia:
Hi there,

Have any basic networking Skills ? Reliable internet for video sessions ? The hangouts will resume fully sometime in the mid of next month. So you have enough time for secure a good internet access if you don't have one now.

nope am just a beginner.

OK. You need to do some self study first before you dive into Cisco Technologies. I highly recommend you do some surfing online for videos on Network Fundamentals. You'll find tons of them on Youtube and as well as text books you can download and study.

NOBLEDANDY:


nope am just a beginner.

DanIndia:
Hi Larry,

Have any basic networking Skills ? Reliable internet for video sessions ? The hangouts will resume fully sometime in the mid of next month. So you have enough time for secure a good internet access if you don't have one now.

Have gotten the basic knwledge abt networking bfre,I did it some time ago buh didn't get tru wt it!so when next month are we starting?

@DanIndia Do you really have cisco experience? or you are learning for your CCNA cert.

im asking this because ; u are acting like a teacher here. lol rather than a student. - I am sure this group u guys set up is supposed to be a study group but hmmmm -

Just passing by though! ..... OVER TO YOU!

Did you even read my Initial Post ? If not, did you even read the thread subject ? Noone acted like teacher here... and yes I have Cisco Experience...years of it to be precise and it is a study group. And as per learning for Cisco cert, I consider everyday and every means a learning process. By sharing what I know, I learn.
u are acting like a teacher here. lol rather than a student
Can you please point an Instance where I acted like a teacher ?

Mellorix:
@DanIndia Do you really have cisco experience? or you are learning for your CCNA cert.

im asking this because ; u are acting like a teacher here. lol rather than a student. - I am sure this group u guys set up is supposed to be a study group but hmmmm -

Just passing by though! ..... OVER TO YOU!

We are starting precisely on the 15th and If you can lay your hands on an E-copy of the CCNA 200-120 Official Cert Guide by Cisco Press, that'll be very useful as we'll be using that topic wise. We might move on a fast pace so I'd be glad if you can do some catching up later after the sessions, as the folks and I planning to conduct the sessions are heading for CCNP. We're just doing a CCNA brush-up.

issylarry:


Have gotten the basic knwledge abt networking bfre,I did it some time ago buh didn't get tru wt it!so when next month are we starting?