High Availability Infrastructures for SMBs


Thank you for showing up for my presentation this morning. When I got up, I figured I was nervous for not having anybody show up for mine anon really nervous before I begin, I’m just gon na say a few words in French and I’ll come back to English. Bozos representing kazoo a lot I saw since the opening tip Olympian continually qubcois travel technologies and ferment de facto Blanc continue. Gon na load, nerves and Canada’s American is repelling pollution. Depan Connie, so YouTube live OHS, webinar, swimmer presentation, a glance upon a CVD casts yong-jin period.

The question offenders so long Geneva, papazilla Dan along the boundary line, a mainland okay, so back to our regularly scheduled show my name is Anne incas, oh and I’ll, be presenting high availability infrastructures or say that real fresh three times on your Frenchmen for SMBs. First of all about me, I’m a an electrical engineer and I’ve been working over 20 years with in telecommunications, various OSS, like Windows, UNIX yeah, I’m old enough to remember. Unix Linux. I’Ve dealt with many different types of companies like privately owned government entities, guys state-owned entities, and I also authored the latest version of NTC n, a course material which new users and new trainers can use if they wish to. And my wife and I, which is sitting right there, usually Oh, am I going to sleep in the couch tonight.

We founded the company called sounded service professional on its gasps young. We specialize in educational services. She deals with the younger crowd, she’s, a special ed teacher, and I deal with the older crowd the adults. So I give micro tech courses, I’m certified, CNA, sorry and I’ve just been certified wireless engineer. I do French, English and I’ve also introduced.

Two courses entered into tcp/ip in introduction to mikrotik routers. I’Ve done this because I’ve had students with little or no experience show up in my classroom, and I mean if you ever seen a deer looking in the headlights of the SUV running towards it. That’S what I was seeing so I’m hoping that these two extra courses will have the guys get a better grade when they do the certification exams. Okay, today we will see what high availability means and how to use mikrotik equipment to build a robust and flexible network. We’Ll begin with some theoretical aspect: we’ll touch about the concepts of HJ and by the way I will say H a as a French winner will be much simpler for me in my tongue, we’ll see also what the major causes of downtime are and we’ll also see A general example of what redundancy should look like for us in our IT rooms, then we’ll delve deeper in the technologies that I’ll be talking about and we’ll have a pre-recorded demo.

I am not so masochistic that I will do a live demo today. Those never work. I’Ve classified my presentation as intermediate to get the full of today’s presentation. It would be good that you have a good understanding, router OS and the general concepts of networking okay to the meat and potatoes now wikipedia defines high availability as a characteristic or system which describes the duration length of time for which the system is operational. I define H a as uptime, I keep my job time in happy Bastien and isn’t that right when people talk about a chain, we also it’s always attached to a percentage score, and we always also hear about the dreaded nines that follow that score.

So basically, we all know for those who have had the problems of dealing with a chain. Those nine means those nines mean big bucks. So what do they mean these lines? At any rate, so I’ve just grabbed this from the internet and, as you can see, the first row is 1/9 and the last row is nine nines. Now I don’t know about you guys, but I’m not so fast that I can fix a bug in 32 millisecond.

Even if it’s once a year now, obviously these scores are totally biased and unrealistic because we haven’t factored in scheduled downtime. So when we create an infrastructure, we go to our client either it be our boss or the client of our company, and we will have to negotiate downtime, which in turn means that we’ll have a realistic, uptime score. What are some of the down times? What causes summertime times? First of all, we have hardware failure who has never had a router go poof, even if it is mikrotik, router right network failure.

I mean you’ve, probably put in 1 instead of 0. The comments editor we’ve all done, that human errors disturb overload electrical supply just show of hands here who was ever, who has ever caused an outage by mistake, of course, in their networks? Oh yeah, I’m not alone rule ok, saq Thursday afternoon 5 o’clock. I brought it down to its knees, I mean for those are not from Quebec. That’S the Quebec liquor board like a LCBO and so on.

My bus was good to me because I I came up and I told him and he gave it to me, you had the right for one mistake a year and you just did it and I for sure thought, like my buddies and from the DBAs would kill Me, but they got so much overpaid overtime. They actually liked me, okay, so now hey. What would it look like in a dream scenario? Where money was no object? We have double core routers.

We have double links going to switches, ideally and our my server buddies with. We agree for dual nicht servers. We have two ISPs wisps, of course. Obviously different ISPs don’t want to go to vader, Tron or Rogers, and if they go down you’re, both your links go down right, you’ll have dual electrical supplies and everything you have trunks. You have VLANs hanging over the place about the only place that you won’t double over are the Knicks and your PCs and printers and phone I mean there are limits to how paranoid you you want to be in life, okay, topology.

So how do we get there? As we’ve seen in the previous slide, we have duplicate Hardware, duplicate lanes, typical duplicate electrical supplies, electrical circuits, 120 or 240 power supplies inside the routers. Now is mr

Street guy here I know the gentleman form a critic, I believe, will have a warrant that on his robberies this afternoon, about duo power supplies within your router’s. But doubling everything over is not the end of it per se. You won’t have to have proper configurations like VLANs, ERRP and other various optimization which I’ll touch upon in the coming, slides, ah having fun now technology’s legal disclaimer.

What I’m going to show you I made for the purposes of this demonstration, it’s small it’s fast and it worked. I was my main goal and you’ll probably will prompt we’re a big group here who could probably have our own ideas and suggestion and how to make what I’m going to present like I said this is just for demonstration purposes. First of all, I’m going to talk about flexibility. Merriam-Webster defines flexibility as characterized by ready capability to adapt to new different or changing requirements. What does that mean?

You want your network to adapt quickly to organizational growth, demands and changes, while at the same time minimizing those costs. So what do I mean by that? We all started one day, creating our network and we were pretty much wet behind the years we had went subnet B, it switch or mostly most likely hub in my days, run router Internet access and woohoo. We were king of the mountain people by the way all the colors are different departments, and I mean we could go and Internet surf whatever we weren’t supposed to serve and it worked right until one day design confronts reality. Your boss comes over, he says we’re expanding.

I want the yellow group to go on a second floor, but the one guy from the yellow group has to Sanger first floor, but he still has to be plugin. These networked are great. Another cable to pull and the one guy from green and red are going upstairs, but they still need to be plugged downstairs own great two, more camels, oh and by the way, they’re, yellow they’re red guys are playing so much halo that they’re really slowing down the Network, so I want to isolate these guys. I don’t know what you just fire them right off: the bat okay. So what do we do?

We implement VLANs, so VLANs are a layer to technology that we used to partition networks into virtual broadcast domains no later well. It’S like putting two hubs: a to switch side by side layer, three yeah! Well, I always expecting a question. Isn’T like layer three supposed to cut broadcast domains? Yes, it is, but even if you haven’t managed switch and you do two VLANs, you won’t hear the VLAN ones broadcast over VLAN torrent.

It will just cut right there, okay, so my second design having been burnt by reality after meeting that as design would look like this. You have one switch on both floors. You have just one cable being pulled to each floor, simple and then you have your VLANs. Being propagated all over the place, so no matter where your boss decides to sit their employees, it’s only a matter of going back to your router or switch a little little commands will apply. Okay and you’re done now, one when you want to configure VLAN.

How would you go about doing that? First step and again, this is my suggestion, because I know you can do with switch chips, but I’m a lazy bastard a lot like bridge interfaces. So the first step that I like to do, I create a bridge interface, which I call BR trunk. Well, you may have guessed bridge trunk by the way when I give the courses when i program my routers, I emphasize one thing amongst others: make meaningful names, because if you put trunk or bridge blah blah blah, it won’t be. It won’t speak to you so to me.

For me, this means it’s a bridge for trucks. Second step, you do you create all your V line interfaces, one by one and you’ve notice. I’Ve created two Y X Y Z under the this bridge interface. Third step I mean, if you want trunking, obviously need at least minimally one physical interface, so I’ve programmed one interface under the bridge which will be the trunk. So that’s the trunk recipe second step.

I want to create a regular VLANs. Not before I go do I have one more step slide to show you afterwards, so just wait first, step for to make a reveal and work in my recipe. Again, you create another bridge interface, which I called in this example. Br VLAN X, the X, meaning one two three, whatever you want second step, I take this VLAN interface rule, which was already configured from the previous step, and I assign it here and also you’ve noticed. I’Ve are giving an IP address now, you’re going to say this is where people would go well.

Where do I put my PC? Well, this is just the two steps you need to do to make your VLANs work now, if you want to plug a PC. This is the third slide you will, you will create an access port. This is already there. We saw just grab a port, assign it into that bridge, and now you have created your first access point.

So previous slides, we saw we had one trunk port. Now we have one access port. So how do we do it? I like CLI, is probably a lot of you do also. I mean this will speak volume again right as a resume one bridge interface to host them all and one bridge interface per villain.

So, in the end, it will look like this: the bridge we have three bridges, one for the trunks and one per V line. Now, when you create your V line interfaces, what you’ll need to do is this is at CLI commands. Now I want to bring your attention to the the blue circle right here. When you look at the mikrotik web pages and the forms and the wikis they always use in the world’m. It’S not Microtech is like the people who feed the page.

They always use ether v as an example to just demonstrate how the concept is done. Well, that’s, okay, knit worse, but don’t what does the recipe I propose right this morning? Is you don’t want to create an infrastructure where your ether 5 is hard coded as a trunk, and if you want to change something around, you have to go through your entire configuration and start wondering. Am I missing something if you use bridge interfaces, if you want to just take five and put three in its place, it’s just a matter of reassigning and that’s it. Three becomes trunk and five becomes whatever else you want it to be.

That’S it’s a little heads up. I just wanted to bring to your attention next step. Where am i okay, the trunk party, either now we’ve seen, we’ve done the Britten, the bridges we’ve done, the VLANs. Now we’re going to start binding stuff together because we have the ingredients where we still need to make the cake now to make your trunk ports you’re, going to take one or more physical interfaces and bind it to your bridge interface, the VLANs in the access port. Again, I must emphasize the first step is without this: no violence.

So you, you bind your VLAN interface to that bridge interface and, lastly, you will take one or more physical interfaces in bridgette to that one bridge, VLAN interface, and now you have just created one or many access ports. So so what does it look when you look at this CLI? Well, basically it so this example. What it demonstrates is you have a trunk and you have one access port and whatever traffic comes here will propagate through that bridge interface and if it’s meant for some other router across a trunk, well just flow over to this VLAN interface and finishing touches. Well, yes, you’ll need to do some IP addressing for your bridge VLANs.

Now I emphasize here trunks being layer to don’t need. Obviously, an IP address you’ll create your DHCP servers again on your VLAN bridges, again trunks being the layer to don’t need servers there and you’ll finish off with a little inky dinky parameters that you really need to make the whole pudding stand together. You configure DNS and TP identity, etc and so forth. Ok, yes! So now we’ve seen and pretty maybe this a lot for a lot of you.

This was yeah. I’M doing that and I get what you’re I see where you’re going this step is. It is interesting now you have to optimize STP priority, so how this is done. You will open your, for example, my example here your BR VLAN bridge interface and you’ll notice. It has a priority setting here by default, it’s 8000 hex.

So what I’ve done in my? If you remember the diagram in the beginning, we had two routers master, backup the master you’ll, give it a one thousand value in hex and the backup a thousand and one yes. This is a may be like going going against our grain because you reuse to bigger number bigger Priya pry already, but that’s not the case for steeping our STP. So again, if I repeat, master a thousand backup 1001 and you keep the STP priority for the other devices at 8,000, their default, so why would we want to do that? Well, let me tell you about the Forgotten bridge, okay, so much for my Vincent Price, imitation.

Okay, this, I kid you not. It happened to a friend of mine and it was a very big mystery for him, the scooby-doo mystery machine. He had finished upgrading his entire network. I mean what was modern back then, but the thing is, he had forgotten one little inky dinky two-port bridge and it was left behind in a photocopying machine. So whenever somebody would come around and wait for his hundred page job to finish printing, what does it use?

What does the usual ad D guy? Do? Click, ok, light turns on huge weights around the lights, boring click it off again. He would turn the bridge on and off again. So what do you think happens?

It forced the entire network to reconverge. Every time add a layer to land level and then the boss would storm into his office and your network sucks and it’s not stable, start sturdy and worst of it is that particular bridge. I had a higher priority than the all the other switches default priority. So, of course, whenever the state of that bridge would change, the entire network would converge. Luckily, he found the culprit and he smashed it.

We had 20 pound sledgehammer. I remember that day perfectly it was beautiful and he kept his job. Ok, next, optimization, which you would want to look into again when you open one of these, the physical parts, two parameters that you’ll see, are edge and point-to-point. So when, when you know that this particular port ether for, for example, plugs at PC a printer, what not and it’s not going to be a switch or a hub – I suggest put it to yes. So you tell that the at that port is an access port and if it’s a trunk towards another switch leave it to know.

So what that does it’s going to shave off a few second from the negotiate, your arse TP negotiation and now because of that nerves, I’m losing it yeah. Well, remember: discarding learning and forwarding States. So when you put edge yes, it goes right away forwarding and you shave off about maybe like 10 seconds of waiting waiting time before you traffic and start flowing again in Cisco, yeah and then point-to-point. If it’s a full duplex link, put it yes and half duplex link, leave it at no and again you’ll save a few seconds from the convergence states. Okay, so, obviously when why do we do this?

We want to speed up convergence and also – and this also I emphasize a lot when students come to my classes. You want to be able, as sis and mints as designers, to be able to predict and say exactly how your network is going to react. I mean what’s the use of creating master backup core routers when it’s your third floor switch, that’s our STP master. I mean it’s kind of useless. You don’t want, like a network, don’t want to create a network.

That’S totally surprised, total surprise for you, water best invention in the world after beer, redundancy, heirloom, Webster defines redundancy as a part in a machine system, etc that has the same function as another part and that exists so that the entire machine system etc will not fail. If the main part does so well, the question is the answers are pretty. Obviously I mean you don’t want downtime, because you want, you, don’t want the loss of productivity that comes with it with it. You don’t want a loss of revenue and, if you’re a service provider and then solve during an ISP, what not usually those dreaded SLA s probably carry a lot of cash penalties for you guys and we’re all working very hard to make our money. We don’t want to give them our money to customers, because our infrastructure wasn’t as sturdy as it should have been.

So if we remember our second there when we left off with VLANs, this is what our second design looked like. Until once more design meets reality, we lose the rooters, so we’re begging and groveling and bootlicking to not use an a word. So we get to keep our job provided that this incident does not reoccur ever again. So now we’ve learned we started from doing lay flat Network. We learn about VLANs and now we’ve learned about redundancy.

So what are we going to do about it? We’Re going to use Hardware, redundancy link, redundancy and we’re gon na touch on vrrp. So now that we’re bigger and better and stronger, this is what our third design would look like. So now what happens is so we lose up to should our network experience up to two outages, which is the case here. Our traffic would still flow so traffic out bounds like this inbound, while I’ve simplified the Green Line, but you kind of guess it would go through here and back here.

So we could lose two links, two outages and you still get to keep your job. We can lose an entire router and you still get to keep your job, because what you did actually works. So it’s hardware, then you see enough new. We need to we’re going to look at vrrp now, so this protocol stand with the acronym stands for virtual router. Redundancy protocol and it’s a protocol that assigns a responsibility of a virtual router to one physical router inside a group of one of two or more of this thing can have a group of one and it shares also the control of a virtual IP to that active Router inside that group, so why should we use the RRP or, as we all own telecommunications, our PCs printers need a default gateway to just get outside to another subnet, so the RRP ensures the availability of our default gateway.

So this is how I virtual I visualized vrrp, so the yellow box is the VR one virtual router. So this is what virtually exists now that we’ve used to physical routers, the green and the orange ones in the vrrp mode. So if you look and if you, if you look back at a few slides, we’ve seen these three guys here so we have VLAN alpha yeah. This example sort of I’ll be making sure that VLAN alpha is redundant, be using the RRP. So we have our VLAN interface, which is bound to the bridge of bridge VLAN alpha and we have a physical interface here.

That’S it’s bound here same thing across on the other router. The only new thing that we added on this cake were the candles. Okay, the RRP interface we have – and please notice also when I talk about meaningful names, vrrp VR, VLAN alpha. So I know that’s it’s a vrrp interface interface for the bridge that controls via an alpha same thing across here, and I’ve also said that you have a virtual IP that is going to be shared amongst the two or three or X, and it happens to be Here now, you’ll notice via IP address alpha and alpha, it’s a yeah, it’s and then you have IP B and C so do take notice three different IP addresses. So how do we do this?

Well, very simply, it’s just another. Basically another interface. That will add. First of all, ask yourself the question. My first router isn’t going to be the master router or the backup router, and do you want the routers to preempt now what is preemption if my left-hand router is May and goes down, while my right hand, router goes!

Okay, I’m gon na take over because you’re, obviously on sick leave. Now, when my left hand, router comes back up, if he’s not pre-empting he’s gon na say well, he stays there you’re doing a good job guy. But if I am pre-empting here, he’s gon na say give me back my toys. I want to play with them now it’s over, so this guy’s got to go back into a waiting mode for my body to be sick again. So on my green router.

What I’ve done is I’ve created two vrrp interfaces, because I have two VLANs that need to be protected default priority being a hundred I’ve increased the priority here so now. Just because of this, I can see that this is going to be my master router for the both VLANs and they are ideal well by default. Er ID is always one, so don’t forget to search change your VR ID otherwise you’re gon na wonder why nothing’s being taken, nothing works for you and on the second router, but the full preemption is always on, so you want to take it off this guy is The backup router it doesn’t need to preempt, because then you turn away. It’S gon na have a lower priority anyway, so it’s kind of useless. So pardon me if there’s no delay with me like delay after a certain amount of time.

No, it’s either either. You can’t that there’s no parameter like wait for a certain amount of milliseconds or seconds. I mean it’s either on or off, but on yeah. The question is, can i delay the preempt and the answer is no, because there’s no value for that, but comes to mind a little thing. I want to bring out you’ve already experienced one outage, because your main router went down right so sometimes and some sensitive networks.

Some people won’t like to do is disable preamp totally because when the other one comes back up, you don’t want a second outage to hit you in that face. So some of this side, while I’m gon na wait that till 5 o’clock, when all my users are gone in the the office, is closed to just switch the the priority over back to the main router okay. So next, okay, so we’ve done basically now we right here, we’ve done the interfaces, but there is no IP attached to it. This is where we are so we’re gon na make an IP address using well. First of all, obviously a free address.

We don’t want duplicate IP address and the particularity of this beast is it’s a / 32 mask. So yes, I mean, if you don’t believe me, go on my critics. Webpage grab the RFI, the RFC, sorry and it’s a slash 32 and you will assign it if I go back a few slides right here. So let’s say your IP address is 1 1. 1 1.

Well, both of these routers will have 1, 1. 1. 1/32. The low the only difference now is if this is a progression of your design, as we’ve seen from the beginning, you’re going from one to two routers, your original routers probably were one and two well now. Obviously, you’re gon na this guy is going to become point.

Two and he’s going to become point three way: is that clear for everybody about the the little room bound with the IP addresses and little piece of also, this little note here never have identical ethical, real and shared IP addresses. This is what we called a honor mode and not all routers support owner mode. Mikrotik does not so again. Your vrrp interfaces will have, for example, dot one both of them, and the physical interfaces will have dot two and dot three now optimization. Well, it’s not really optimization or shorter may be rewrote the terminology here, but when you design your vrrp one will be master for all your VLANs one will be backup for Allah VLANs and if the master fails, the backup just grabs the default gateways.

We in a matter of one second and that’s over, and if your backup fails well, who knows it wasn’t doing anything anyways now I know some of you are probably going, but what about load sharing? Keep that caution in mind I’ll get back to it by the end of the presentation, because there would be a better way of doing this and I’ll explain why we won’t be doing this right away. Okay, various ok! Now we have a layer, 2 and layer, 3. Router network and works fine you’ll obviously want to be doing some dynamic routing.

Nobody in his right mind likes to do static routing, especially if you’re beyond 3 routers. What I suggest in this case you open your OSPF interfaces and you assign them. You change their priorities. Now, if you all remember the default priority for interfaces is 1. What I’ve!

What I’m suggesting today is your designated router or your dr put it down 100. So, whenever your routers come up, he’s gon na be the dr. Your b, dr give it 50 and your members give it give them 0. There are only members, I don’t think they need to participate in the OSPF election as long as they get the routes. You’Re gon na be ok, now again, you’re gon na do this to make everything predictable.

You know who your dr is, who you’ll be DRS and you your members are but OSPF being the non egotistical child that it is. I mean if the dr goes down. Bdr says: ok, i’m a nude beach. I mean –, you, dr now, and when the old dr comes back up well, the new, the backup router will stay. They are so that’s something that that’s always PF.

You can’t change it, but again, if you’re gon na you might not want to cause a second outage. So you might want to leave that for the end of the day, when all your users, a gun and especially the boss, is not looking at you one thing and it’s useful here, don’t see if lower your hello intervals and you’re dead intervals to make convergence faster, But, of course, the hit on that is a wee bit more OSPF traffic. So it’s up to you individually, to evaluate what’s good and what’s not for your particular case, pitfalls, Anya, maybe sharing personal experience when you start playing doing high-availability, I started off with fresh routers. No default config, nothing. So what did I do plug everything start configuration bad idea, Joe, because spanning three loops traffic goes through the ceiling, CPU chugs down and I’m not a happy camper.

So when you start playing with setups like that or any setups, where you have loop possible loops, just configure your routers and plug a plug afterwards, ok, micro tech, guys here nobody from mikrotik, I had a wish for them. Jeez, no red shirts! Well, I’m gon na say it anyways, maybe somebody’s gon na squeal on me. I would like for MSTP support godless MSTP support. Why?

Okay, because remember last time early, I told you the way, I’m doing H a right. Now, it’s not as optimal as it could be. Let me go back to my: the joke is lost on them now, they’re not even present. This is the demo that I used in my in my lab now without MSTP support or for those guys who know cisco, / VLAN spanning tree protocol what’s happening here here is i’m naming this guy master layer 2 and master 4 vrrp, and this guy is backup. So what’s happening, this got this port here is blocking, so all traffic on this switch is going through here to this router to the internet or what a server farm whatever it’s coming back now, if we had MSTP support, let’s say we have four lands: the odd-numbered Lands and the even-numbered lands.

Well, you could say this: guy is master for the even-numbered lands and vrrp. Obviously, and this guy is master for the odd and so whatever traffic that goes on two and four would go here. One in three would go here and back and at least now, both routers would get half the load and have to work and all that pay and all the links would not be topped at a gig or a hundred or whatever configuration you have any. And if you have an outage now, that’s a cool thing. Instead of let’s say you lose your main round because that’s the the problem here, instead of all your four VLANs being hit at once, you will only get half the VLANs being hit.

Unfortunately, without MSTP support. We’Re not going to be able to get that so again and yeah. That was maybe a middle marketing speech for micro tip, but again they’re not here for those who want to migrate from Cisco. It would be so much easier for us to do right. So I’m at the part of the un — live demo, so I record – and I hope you like asterisk music – that’s all there is ok there.

First, this is my setup two demos at first, the first demo I’m going to simulate this guy going offline. Second video. This gun is going offline, all done. Okay, I just for information. Before I proceed, I had a right here.

I I used another router with that watch just to be able to ping all the addresses that I had to ping so feel like a preacher. Okay, so you have all the addresses. I finished that by dot one are my vrrp interfaces. The dot two are the back up, the main core router and the dot threes are the backup, routers, okay, and here is the IP address for mikrotik, and this I can’t remember what it was. So it’s pretty cool the router.

The VoIP call, which was on hold only lasts, maybe like five seconds total and as we saw the ping towards Margaret six website had to suffer the same ill effects and now the second router. I guess you can pretty much guess. What’S gon na happen being the core router backup? Okay, so final words, what we’ve seen in this presentation was the flexibility of using VLANs, the redundancy of doubling links and hardware, and by using vrrp and all of this put together, we I demonstrated how you can achieve higher availability of your infrastructure and also make you Aware certain things like optimization and steps to follow not to crash your hardware. I have some time left for questions.

If you have any okay, she sure the the first one its second one, the first one her last one. That’S all, let’s see one action hold on. Is it the this one here mmm-hmm? I missed a few of you. Okay, so if I’m understanding carry on our repeat your question, you would suggest using scripting to just switch priorities from one router to the other.

Is my current like going like this okay because you’re still having problems ending all your your question but you’re saying like disabling this link here? Well again, because what you’re suggesting is you’re using automation to this to disengage? Or this like, we move that link. What I’m suggesting is I’m from a an era. Is I don’t want to create all of my tools and by using MSTP or syscalls PV STP as it was?

You still had to have that link in the middle because, if you do lose here and here without that cross link well, your traffic is not going to go out to the Internet. Yeah because I mean mikrotik is extremely flexible and what you can do so stuff. That’S not included in our router OS, yes, you can work around it or band-aid it, but I’m I know maybe I’m too lazy and I still yet have to learn scripting. So I think I’d rather have the full protocol support and having to patch my way through it. Well, thank you a good comment and, as I see and this setup here, okay, the master/slave for layer, 2 layer 3.

Well what I’ve done! That’S like, I said the. If I go back to the layer 2, if you layered 2 bridges, if you’re talking about well the question first of all, is how do I set up my interfaces just so? Okay, so the trunks, if we remember, we had two parameters that we had to play. The edge trunk with here would be no because it’s a trunk port or another switch, so it has to go through the RSTP steps of listening.

There were three steps on my memories shop this morning, but there were three steps that you had to go through before the port. We start forwarding so this these are I with you. This would mark as edge equal. No, you could put Auto and you would start like doing a little extra mumbo-jumbo, but in the effort of speeding up your infrastructure, I always like to hard-code Maya my configurations and answers so edge and point-to-point well edge. No point-to-point!

Yes, because it’s full duplex and the bridges here would have a priority of a thousand hexadecimal, and here we will have a thousand in one hex. It does that answer your question: no mm-hmm! Oh okay! Okay, that, okay, okay! So the question here is: how do we set up all the Ethernet ports more at a hardware level?

Okay? Well, I found the ports were always on the same VLAN. You could have slave master relationship. I personally don’t like it because well for my environment, I don’t need the performance boost as much as I need a flexibility. So in that setup I would just leave all math all ports, no master, no slaves and just bind them into a bridge to the CPU yeah.

It’S switch. It will burn through CPU. I know it’s not optimal and again, like I said this is my recipe for this demo, but it’s always good to look at what you need and if your environment is very, very stable and static, you could go about the use you were going with the a Slave master setup, yes, I’m and I guess I’ll start questioning this also Beijing. No, mrs

Otter apollomon step for the white which proposed a parallel, switched ship gaps, but finding plug Lord Rama in blur glue, vassilis empty Sierra Dan Danko emoted to provide suit even on BG software, ooh sweetie hardware become encased hood semantic, esta camiseta hood master port. We serve a marshy, mainly these eat up, sometimes my soap opera, Oreo cookie la fecha, low fat low.

So you still accept objective, mag catch the rabbit less, which chip to push water and I set up master slave 62 gene and that definitely Marisa song, Jae Suk’s love longer gives a burst past. I had no question okay. Just to make sure I wanted to tell you a question you’re asking: if there’s something in the logs when and vrrp event happens, yeah, obviously well, I guess I could. Probably if you, if I just look quickly la system logging know my memory is terrible. That’S why I’m so good at documenting, but if there’s the RRP there is vrrp in logging.

Okay, so you would have something there, but also you probably want something else to log the event like the dude or any other network, don’t manager and keep a history of, because I don’t suggest using a router as your main information database right. So you might want to use a SNMP, but what’s the other protocol I 123 at non syslog. Thank you now. If I had a pen, I would throw you one yeah. Well I mean processing and finding information.

I’M not a big deal. Really. I mean just a new little UDP packet, that’s being sent so yeah. Sorry, if you want to keep a history of the events, I would definitely go into syslog away. I mean oh yeah yeah.

Definitely the you can configure syslog and whatever you’ve done in your system. Logging subsystem, you can always add an extra action or yeah and send it to your server, which is a different way to go to oh yeah, I mean anything I can happen. I mean the worst that you could do is activate debug everything and Senators syslog. I mean for those who are on the in the Cisco world. I have once had a friend of mine do debug all, and then we had to reboot the server period.

I mean the switch, I mean a 4340 500 switch. I mean she really killed it, but you could probably do the same in micro, tech and just do a target, your your events and send them to syslog. If it mean it doesn’t work with non mikrotik yeah with non Cisco. No, I mostly play with Cisco and mikrotik. I know I remember, like I said earlier, my Cisco.

They had pv STP and that was peachy to use but other non Cisco. No I’ve never had experience questions guests, young, that’s it! Well. I look. Thank you very much for your attention.

My throat is raw and dead and okay. Thank You microchip for your first mom in Canada. It’S a soap we’ll have more and enjoy your hair. The rest of the show. Thank you guys.

Gals

You May Also Like