 So the other day I did a video called the Unify Controller 510.15 stable has been released and now that can be followed up with 510.16 and 510.17 stable has been released. What's going on Unify is what people are asking, but you're not asking the right question. Why aren't more of you testing these when they're in beta and this is where I'm not ranting about it and I'm not saying any company is ever perfect, but there is a lot of scenarios out there. There's a lot of large unified deployments. We have some of these large unified deployments set up. There's a few things that you need to know about this. Unify is based on the MongoDB system. That is kind of the backend that drives the pretty web interface that we all love. And one of the things about MongoDB, it's a little bit, I don't know, we're going to say tricky when it comes to database support. I wish they did not build it up MongoDB. I don't have control over that. But that being said, that's why there are things like this, the Unify high density WLAN scenario guide. And it walks you through a lot of planning settings for things that are important so you can figure out how to do some capacity planning. There's other guides on how to finally tune the Unify controller for that. And I wanted to bring that up as well because that's sometimes where the first problem starts is you just use a stock controller and then deployed 400 access points on it. And then you go, well, the upgrade didn't work. And I was trying to run 400 access points on my cloud key. You have to properly spec any of these and read about the product. Just because the product is inexpensive and accessible doesn't mean it automatically just works at large scale out of the box with the default configs. And you can crucify them if you want for that. But the fact that they provide documentation means you should probably take the time to RTFM before you go online and start yelling about it, which I just see so many people do. So I want to walk you through though, our upgrade process for Unify. Because when I say I didn't have problems, one, I don't mind testing out their newer software because I understand how the backend works. I understand how to back things up properly. So I understand how to do these deployments. Also, if you want a product to be better yelling about it on the internet doesn't help putting a concise scenario where here's the settings I had, here are the problems I had and posting it in one of their forums is a bug report. So we can all come to a better conclusion as to how to solve it. That does work. That's what forces all these release updates. So they were able to find problems that people did report. So reporting problems does work. This is a lot of how community development projects work. And it's hard. There's a lot to developing and engineering. QA testing is really hard. They rely on a lot of people who don't mind pressing the beta button like me and also don't mind filling out proper reports because they can't predict every scenario. And then, of course, they expected you to have read guides. And a lot of people I know because we deal with this sometimes have not taken any time at all to read any of the guides or change any of the config settings as suggested by Unify for these larger deployments. And then they have problems when they do the upgrades because, well, they are not running a finely tuned machine. So how do we do the upgrades yourself? So a couple of things. Go over your settings. I'm going to go over here to maintenance. Before I run an upgrade, I run through and compact the database. Confirm. What you want to make sure is that the database doesn't have any cruft in it, any extra stuff in there. They have this tool here for a reason. So I do go in here once a while, run the compact on the database. I want to make sure that that's fine. Also, double check your auto backups. Do you have a backup? Do you have the auto backup turned on? I hope so. This is arbitrarily easy to turn on. We back up our controller every hour. And then this file is actually taken off of the Unify server and put to our pool of backed up files. The reason I do it that way is because it's not the OS I need to ever worry about backing up with this. I only really care about that restore file because Unifies are arbitrarily easy to reload with the restore files. And when I do the backups, I also only care. This is me about the data retention days of being settings only. I don't really care about all the data because it makes the backups that much larger to restore. Now, if you want to back up all the data, you can, you can back up the VM. Maybe you do want to back it up. But for the most part, the logs, if I'm in a disaster situation, that's not what I was worried about. So we only do a full log backup once a week. And to me, that seems adequate because the logging is important when I'm troubleshooting. But from the day to day, having every detailed log stuffed into the system seems overkill. I only do it when we're doing testing. Once we have a site deployed, we can thin the logging out because, well, it takes a lot of extra bandwidth and traffic to manage that. And we manage a lot of sites on this controller for all of our clients. So to have an extended amount of logging in here seems ridiculous and overkill. So once I know those things are done, and like I said, we've already followed all the guides, we have these set up, we have our controller configured with some extra memory to accommodate the fact that we host so many sites on it. We follow the guides, not anything secret there, you can find videos on it. I have not done any specifically, but the guides are really simple. Short answer is we added memory added the number of connections on there. Oh, actually, we tripled the amount of memory it can take. And we added the amount of connections that can take inbound. And that's it. But before we do the updates, the easiest thing to do, I'm running mine in a virtual controller. Now granted, we're hosting it in house. So this is easy because I'm using XCP NG in house. Another way to do this is if you were on digital ocean, for example, you can use their snapshot system, you can snapshot beforehand. The nice thing about snapshotting VMs and many different platforms support this, you have a nice solid, this is where it was point. So we just go over here to snapshot, it's better to be shut down. Someone's going to point that out. Yeah, you're right. You should shut down the controller when you're doing a snapshot. So you get a snapshot of a off state. But I didn't, and I've restored from these and it's worked, but mileage may vary. I know the safer way to do it is as stated, have the controller in an off state when you do the snapshot, that way the database connections are closed. Once we do that, we head on over to the unified site and download the dev package. Now I know you can install this so it will work with like an app to get update. My preference has just been to do it manually because it's not like I'm doing it that often unless you were doing it every time there was an update release on here. But like I said, this has avoided any problems for us doing it this way. Let me copy the URL, W get or curl whichever your preference. I'm old school, so I like W give a curl. We'll do the same thing. If you want to curl download, but downloads, grabs the file. I already downloaded and did this update. So there it is. And it's just that's why I added the dot one off one there. But then we just install it manually. D package dash I unify and run the install or let it go through his thing. When I'm done, I restart the server to make sure everything comes back up properly after it's uninstalling and I've verified that it worked in a way we go. And if it doesn't, I have snapshots. And if for some reason my snapshots were to fail, I have those unified backup files that will help get me through any of the problems I had by just going through here and in XC PNG, it's obviously arbitrarily easy. I can just say, okay, it'll roll back my VM right away to the state it was before I ran the update. And then I leave this snapshot here as I test the update, make sure it works as you leave it for a few days. In case there's any problems and I have to do any reverting back, but that's it. It's all we have to do for the updates. But the important part that I want to get through to people is these updates, there is plenty of discussion. So for anything on here, there's people, there's going to be discussions in the forums on them, there's going to cover all the details. So before you just go around and I see people, you know, oh, their stuff is da da da da da. Just go through and read through the articles, read through the forums, post your problems. But don't just post it sucked or it didn't work, post the steps you went through, post if you did any customization to your controller, because maybe that's one of the things you do. So if you're, I've seen so many people complaining that have these large installs, make sure you've properly configured the controller to handle it, because one of the things that happens when you're doing these updates is it has to expand and make database changes for the new version. Well, if you didn't allocate enough resources for that database to expand, either didn't have enough memory on the system or not enough hard drive space for the database to expand, you could end up with a crash problem that isn't necessarily the same. So as what any other person's running into, so you need to get all those details together, then post it and these forums, because if you're posting it non-official forums, the forums themselves for Unify are essentially the official forums where you're going to find that the Unify network engineers are there watching, they care a lot about their product, but posting it other places, I've seen people tweeting, that's not the way to do it. So you want to do it properly, you want to do it in depth and detail, and this goes for really any product. I'm bringing it up for Unify, but this is how any of those products get better. If you have a problem with it, you go to wherever the forums are for that particular product and document your settings, your scenario, how you did the update, and what went wrong, or at least as much detail as you can muster for them, and they sometimes even have guides that they'll say, hey, we need this log or that log or this diagnostic file to kind of get a better in-depth understanding. So hopefully it's helpful. I didn't have any problems with the updates myself. I do know that some of these updates were releases because other people have reported bugs with the migration, but they properly reported them. They said, here is the bugs, here is the scenario that caused that, and that's why Unify says, oh, we didn't think to see it that way, and they had a update release because the bug was genuine, it wasn't a mistake on the user's part, the bug was genuine. They're going to happen. There's a lot of scenarios out there, software-defined networkings, a big complicated beast. These guys are taming it, and I think they do a good job overall. But hey, you know what, if you want to be one of the people that sits back and waits a week, I don't blame you. If you want to be like me and push forward, make sure you're willing to also document if you're going to run the latest versions and try the things the same day they come out. So that's my feelings on it. I still think it's great. I'm still going to continue to update as the new versions come out. I don't have a problem with it, but I'm also willing to document and talk about it and discuss with the engineers and the forums how that works. So hopefully this was helpful and don't be scared of the updates, but feel free to wait a couple days if you're worried about this happening again when there's another major release. This is only something I've seen on the major release updates, some of these problems, but for the minor ones, the subversion updates seem to be pretty pain-free. I haven't seen them, but that's why if you're wondering why there's been a 5, 10, 15, 16, and 17, and our methodology for handling that. Alright, thanks. We can keep the conversation going. And if you want to help the channel out in other ways, we offer affiliate links below, which offer discounts for you and a small cut for us that does help fund this channel. And once again, thanks again for watching this video and see you next time.