Proxmox VE Full Course: Class 16 - High Availability

[Music] so here we are the final episode of my proxmox course and i'm sad and happy at the same time i'm sad to you know see it come to an end i had a lot of fun filming these episodes for you guys but like i mentioned earlier in the series there will be standalone episodes so it's not actually the end it's kind of like a new beginning so in this video we're going to take what we've built in the previous video which was you know setting up a cluster and we're going to take a look at high availability so let's dive in one last time to proxmox all right so let's go ahead and get started with high availability as you already know from the previous video i now have three servers in my cluster and i have each one in their own tab right here in my browser now i didn't decide to set up three servers randomly three servers is required for high availability now i want to make sure you understand that you don't actually have to have three servers to have a cluster but you do have to have three servers in order to have high availability what's the difference if i have two servers in my cluster then that's the minimum that's required in order to live migrate a virtual machine as you saw in the previous video i can right click on a virtual machine and i can migrate that virtual machine to a different host and it's pretty straightforward but again only two servers are required in order to benefit from live migration now simply having a cluster doesn't automatically give you high availability what high availability means is that if a node goes down then the virtual machines are going to automatically start on another node with two servers like i mentioned you automatically have the ability to manually migrate a virtual machine from one host to another with high availability then you're hoping that if something unfortunately happens to one of your proxmox servers that your virtual machines are going to automatically start on another server as you just saw i migrated this virtual machine over to the other host it was on pve one now it's running here on pve two but i did that manually now when it comes to high availability the reason why you need three servers is technically because you need an odd number of servers so hypothetically if you have two servers in your cluster and you enable high availability it might work just fine you might not have any problems at all until you do have a problem and when you do have a problem it's going to be a big problem so let's pretend that i only have two servers here let's just forget that pve 3 even exist at all so if i have high availability set up with just two servers then both servers are essentially going to vote for themselves during the voting process so essentially what happens is when you have a cluster and high availability is enabled there's a heartbeat between all the servers if a server goes down the heartbeat is lost but never fear if you have three servers you have two other servers that can well take up the extra work it's highly recommended that you have three servers because you need what's called quorum in order for the best decision to be made about which host is going to run the virtual machines that are now not running on a host because it went down those virtual machines need to be running on some other node when there's a problem with only two servers your options are very limited there and also with two servers your quorum votes are well unreliable so long story made short when it comes to proxmox if you look at their documentation they recommend at least three servers in order to enable high availability you could have more servers than that i mean it'd be awesome if i had a bunch of servers i mean how cool would that be but if you have access to three servers feel free to set up high availability if you want to but even if you don't decide to set up high availability you should still cluster your servers because you still have the option of migrating servers manually and you just saw me migrate a server manually and i forgot to show it in the previous video but of course you can migrate containers as well i have these two containers here so i'm going to go ahead and migrate this container over to pve 2. and the thing is that all works just fine as an aside you can't actually live migrate a container notice here it's shutting it down so pings will be lost and the container will not be accessible until the migration is complete it's actually going to copy this container over to the other node and you can see that transfer happening right now it actually copied that disk over to the other server and now that container is running on pve-2 and all of this works just fine even without high availability yeah that's a benefit but it's not the only benefit one of the more common reasons why people set up a cluster is because you're able to update a host and not have your virtual machines go down sure your containers might go down but we already know that containers can't be live migrated for example i don't have any updates right now but if i did i could refresh this right here and of course i'll just go ahead and do that it's not going to show any updates i actually just checked before i hit the record button so that's how i know that there's no updates here well let's say that there were updates i could literally right click the host itself and i could do a bulk migrate and manually migrate everything over to another server then once that's done and all these containers and virtual machines have moved over to another host i can install all of the updates reboot it and then i can migrate all the virtual machines back onto this host or at least the ones that were on this host update the second one reboot that one and then repeat that for however many servers that i have and all of that is a benefit of clustering and we haven't even entered into the territory of high availability just yet i just want to make sure that all of you guys are fully aware of why you would want to set up a cluster and where high availability fits in now i'm not recommending that you do this but if i was to go ahead and just you know yank the power cord out of the back of one of these servers if i had high availability set up then the virtual machines will be running on a different host and that's great that's definitely a very cool thing to benefit from high availability takes clustering to the next level now what i'm going to do is actually migrate the storage of a virtual machine actually i'll migrate all of them to the shared storage so right here i have a container and if i click on resources we can see that it's actually using local lvm so what i'm going to do is just shut down everything i'm going to show you how to properly convert everything over to high availability it's not just enabling settings we want to make sure that our virtual machines their storage are on the right storage medium specifically shared storage so i'm going to shut this one down and this one and this one this one as well and i'll shut down this one too i basically want to make sure that nothing is running as a matter of fact to make this process even better i'm going to delete all the virtual machines that i don't plan on using for this particular episode so we have web server one right here i'm going to keep that one we already know that containers can't be live migrated so at this point in the series containers actually don't have any other purpose whatsoever so what i'm going to do is just go ahead and delete this container i'll remove it make that one go away and i'm going to leave the other ones alone and i'll do some more cleanup here let's remove this container i'm just going to make sure that everything is purged it's going to make this very simple and this server right here is just from one of my experiments off camera so honestly i should have deleted that one quite a while ago let's go ahead and kill it and i'll get rid of this one as well so now we're down to just this vm right here web server one if i go to the hardware tab we can see that the virtual hard disk here is actually stored in local lvm now to get the full benefit of high availability and clustering and this is another thing that doesn't require clustering as well it's just definitely something you want to do you want to migrate this particular disk to a different storage medium specifically shared storage so right here we have an option for move disk i'll click on that where do we want to move it to well what i want to do is move it here to pve shared this is the shared storage that i set up for this tutorial series so that's where i'm going to move it and i'll delete the source i want it to only exist on this medium right here let's go ahead and move it and it's happening very quickly i have 10 gig ethernet it's a beautiful thing it's already done actually so what i'm going to do is just wait for this to refresh there's probably a few other scans or checks that it's doing in the background i'll let that finish and i'll be right back as soon as this finishes moving and now it's done you'll notice that the hard disk is now on pve shared so i'll start up the vm you really shouldn't notice a whole lot different here now to be fair this particular vm might be a little slower because shared storage in my case is actually using spinning rust hard drives so yeah there might be a little bit of a loss in performance but it's going to be okay now what i'm going to do is just log in real quick because i don't remember what the ip address was of this host it's 10.10.10.205. so what i'll do is switch over here to a terminal and set up a ping and now i have a continuous ping over to that server now what i'm going to do is live migrate this server back to pve one in the previous video i already showed you how to do that but in that video i didn't actually have this disk on shared storage so it took a minute or two which isn't really bad but it could be a lot faster let's see how fast this is actually going to go let's migrate it we're going to migrate it again to pve one and it's done so i made sure not to edit that part of the video normally i do edit these particular scenes to make them a little bit faster so i didn't actually edit this part of the video up until task ok appeared on the screen so that was pretty quick and it's still pinging let's move it back to the other server and i'm going to watch the pings more closely let's see if one gets dropped now it didn't actually lose a ping but what i did notice is where it says right here 4.7 milliseconds it did actually pause on that for a little bit longer than the others so it there was a little bit of a delay but it didn't actually miss anything so nothing ever went down and of course it's already on the other server so again everything that i just showed you guys is par for the course when it comes to clustering now let's take a look at enabling high availability and have even more fun but the first thing that we want to do is make sure that the virtual machine that we're going to add to high availability is shut down so i'll gracefully shut down this one right here and now that vm is well shut down so basically what i'm going to do is add this virtual machine right here to high availability if you have more than one virtual machine that you want to have enabled for high availability then you'll set that up on each of those but since i only have this one virtual machine now i'm just going to focus on that one i'm going to click on data center i'm going to scroll down right here to ha i'll click add and then i'll add web server one i want it to be started and i'm going to leave max restart and max relocate both at one those are the default values essentially if it needs to restart a virtual machine it's only going to try so many times it defaults to one that should be good enough so i'm just going to go ahead and leave it at one and it'll try a maximum of one time to relocate the server you can increase this if you want to but honestly if you need to increase these to higher numbers then i would think that your environment probably has bigger problems to worry about than high availability anyway i'll click add and we can see that it's added right here now if you recall this virtual machine was shut down but the desired state was to have it started so it automatically started it for us at this point if the node that this virtual machine is running on has a problem is going to restart the virtual machine on another node so what i'm going to do is actually simulate this by logging into my switch i'll do this off camera real quick and i'm actually going to disable the network port that this host is actually using for the heartbeat essentially i'm just going to pull the plug on it so i just disabled the port and any time now this should actually show that it's no longer available in fact i'll click on it let's check the status right here as you can see it's having a little bit of trouble reaching this host and the x already appeared right here so it's definitely down it's still powered on i just disabled the network port for this particular interface so proxmox has no way of actually communicating with it so i just went over to another tab and i'm on the server right here if you recall that server was running on pve 2 and shortly it should actually go to another machine all right so i had to do a little bit of behind the scenes work here but as you can see the virtual machine did in fact restart on pve one the reason why i didn't do that right away is because well actually i disabled the wrong network port i actually disabled the management network on the server not the network that the virtual machine was actually using but as soon as i disabled that other port the correct port it went ahead and restarted the virtual machine over here on pve one because as you can see here pve two is down it's not accessible so we can see that high availability is actually working now what i'm going to do is re-enable the network right now and we'll see this node right here come back online all right so i've re-enabled the ports for pve-2 the switch that i'm using takes a minute or so to reprovision when i make a change like that and it should be online here shortly actually i already see some question marks here so something is happening it looks like it's coming back online right now and there it is we have the green emblem next to it so it's actually back up and running and this server here is happily running on pve one what i'm going to do is just migrate it back over to the other server and it requested the migration we can see that it's actually in the process of moving and it's already here and it's done so as you can see high availability is working it's definitely something that you should take advantage of if you can if you have the appropriate hardware and if you do well why not but anyway we were able to set up high availability in this video we could see that it's clearly working and that's awesome [Music] thank you guys so much for checking out this series i really appreciate it and i mean that i appreciate each and every single one of you guys because you guys helped make this happen for me if it wasn't for you guys i wouldn't be able to do what i'm doing and i'm having so much fun creating these videos for you guys now if you want to help me out even further please share this content with other people you can share it on your favorite social networking group you can put an ad on a billboard okay well don't do that that's expensive but just feel free to share this content with other people and spread the learning that would be awesome now i'll have other proxmox videos on my channel here very soon but for right now that's the end of this particular series i again appreciate you guys thank you so much and i hope you enjoy your day [Music] you

Channel: LearnLinuxTV

Views: 4,970

Keywords: Linux, Tutorial, Howto, Guide, open source, gnu/linux, LearnLinuxTV, proxmox, cluster, high availability, data center, home lab, proxmox ve, proxmox virtual environment, virtual environment, virtualization, proxmox virtual environment tutorial, virtual machines, proxmox tutorial, proxmox ve 7.0, homelab setup, proxmox getting started, proxmox 7 tutorial, virtual machine, home server, proxmox configuration, proxmox 7, virtual server, proxmox ve full course, proxmox ve cluster

Id: 1nEs1ZvGbTM

Length: 18min 48sec (1128 seconds)

Published: Fri Nov 05 2021