 Here you go. It stops spinning. I mean, I think that means we're alive Which means it's episode 54 of the homelab show me and Jay are doing a Q&A Welcome and And I have to mute my laptop apparently That was really on okay anyway Anyways welcome So me and Jay went through and you have a lot of questions all of you listening that was Those of you that went to the homelab.show filled out the Q&A questions We're gonna start running through them get those questions answered if you want a question answered the best way to do it Is go over to the homelab.show and fill out our form where we collect all the questions read through them Even if you just want to stop by and say hi because you don't have any questions We like seeing some feedback on the feedback form You can also tweet at us and Twitter things like that too We try to compile those in there, but the goal is to you know, always keep everybody informed And you do help steer us into some of different ideas and even suggest new technologies We'll be looking at before we dive into this Q&A episode though Let's thank a sponsor of the show and that is Linode and Linode's been a sponsor show since pretty much the beginning And they continue to support us We think it's a great place to host the podcast, which is why we do it So we that's why or when we make the comment that hey this show is literally brought to you if you downloaded it It was brought to you from a Linode server So they was where we host all the infrastructure for the homelab show And it's also a great place to host all of the different projects that we talk about on this show So we like to thank Linode for being a sponsor if you want to get started with the node We have offer code a homelab show to get you started with that I should think it's the homelab show I got to get that right make sure people do it either way It's in the link down below Podcasts on the website and in the YouTube description Keith Tom says it wrong And so thank you for sponsoring and let's get started. Let's get to it So first question we have on here Jay The first question is how old is too old and that was sent in by rp And basically the consensus here is that we're always or not always But we often recommend older equipment because it's cheaper because if you think about it You know, it could have this crazy awesome server that might have cost, you know, several thousands of dollars for a company But then, you know after some time goes by it's, you know outdated to that company It doesn't keep up with their demands And then next thing you know, it's for sale on ebay four hundred and fifty dollars Which is basically next to nothing compared to what it was sold originally for And at what point does that become too old for, you know, too slow for the homelab user and The thing is when you think about servers, they are very often Um You know very powerful machines because they keep up with a lot of users But in the homelab, you're not going to have nearly as many users even A small company will have more users than you So it really takes a lot in my opinion for these things to slow down I think that power usage is a bigger concern because the older ones are super, you know Expensive in power. They're just not very efficient. That's a bigger problem But as long as it has, you know, the specs you're looking for like the amount of ram It supports 64 bit operating systems, which I'm not aware of anyone any of them that we talked about that isn't You know able to support that so I think it's more of a question of the power usage because That's where it's going to cost you a lot of money over time Especially if you live in an area where power is expensive. That's definitely a bigger concern But I would say the in my opinion the r6 10 r7 10, you know 720 Um, I would say they're still relevant. Um, the other thing that's pushing a lot of people or services availability of them Because some of the newer equipment is a little harder to come by Um with some of the supply chain issues So it's occasionally can be tricky to find and some of the other commonly suggested low power devices such as raspberry Pies are less accessible right now as well Right and there's two folds to think about is one the electricity cost to The heating or cooling costs heating really it's heating up your area Maybe you live in a cold climate and that's fine. Maybe you don't and it's very much not fine So there's something else you have to consider is that higher wattage Is dissipated as heat and then depending on where you live, uh, you may have to Reduce the temperature and uh, that's a different problem. So something else to think about it's another cost It's just the cost of the wattage it uses and then the btu's It creates then they need to pump that heat out of your Living space unless you want it that much warmer. So those are all the considerations when you're looking at some of the older equipment It may work perfectly fine. Um, but that may be the bigger concern for you. Mm-hmm So how old is too old is that balance? Hopefully Yeah, you know, we don't really know what you're running I mean you could be someone that's running open stack and you have you know Or at least that's what you want to do Which could really, you know, bring some hardware to its knees because sometimes that could be a very, you know, high Usage platform. I mean, it's just gonna use cpu and ram like crazy Not something I technically recommend But if you're going to be certifying in open stack, you need something to play with there You know, you know, maybe then you could make a case or something that's more powerful But for most of the things that we're doing in the home lab I just don't really feel like we need a crazy expensive new server. We're pretty good with what we could find Yeah, um, the next question about rj45 ends So this is a tricky one because there's a little bit of confusion You'll see rj45 ends labeled whether they're cat 5e cat 6 cat 7 The bigger difference in them electrically, they're not going to be substantially different at all What they are going to have a big difference in though when you talk about the crimp end itself Is the fact that the cable diameters vary by spec So the cat 6 cable, uh, or cat 7 cable is going to be thicker So you have to match them on that but in terms of will they work? Yeah, you can sometimes stuff and Cat 6 and 2 a cat 5e cable It's not ideal, but if it's what you have it shouldn't make really any speed difference because you're not dealing with Much in the way of transmission data We think about it. They're just tiny little pins Connect any cable as long as they have a good connection to that cable You're fine. That's not where there's a major because functionally they're the same design wise They have to snap into the same part whether the cable coming to that point Is a different type doesn't really matter much because they still conform in the plug part So as long as each little of the pins touch the wire, that's not where you get a speed difference on there Also, the person had mentioned they're running about 80 meters I want to know if they should go with something more than cat 6 and I don't see any reason to do that not with an 80 meter run Cat 6 will do if you go with cat 6a specifically you can do 10 gigs at 80 meters with cat 6a So there's not really any need to go for the heavier harder to work with more expensive cables There's not really much of a benefit for any of that for those of you running and for Those of you that are willing to gamble a little bit Not that I'm recommending this but for those you wondering if you have to replace the wire to get faster speeds One of the things I've done if you look at my channel I've got some out of spec cabling videos where I talk about the fact that cable usually works a little better than it's Rated so it will not certify it will not pass But you can get away with and for a home lab that may be perfectly fine It's not a production environment. So if you go can this cat 5 cable in my house Over a short distance transmit more than one gig probably actually so Not a guarantee, but it's something worth testing before you rip and replace it If you're putting a new go with the better stuff, but if you have existing Don't go through the trouble ripping it out until you've taken the time to test to see if it meets your needs Yep, and we had a question about you the key in reference to episode 45 specifically Um wanting to know if we have checked out the only key. I haven't have you checked it out only key No, but let me grab it out of my drawer so I can remember the name of it I have more keys I bought more keys because uh, I have a trust key in my hand. I had to look at it to remember the name of it Um, I'm going to I'm I didn't know if I want to buy the other ones because I remember looking someone We're out of stock and trust key is part of the fight old alliance I I'm going to do a video because I think fido is a solution not any proprietary standard so Key authentication awesome key authentication using fido amazing. This is something ube key supports This is something that other than ube key hardware key authentication can do Companies have been certified with the fido alliance Have gone through security testing to make sure that they've implemented it properly Any well implemented properly implemented fido key is really good for security The challenge you will run into that i'm running into even making the video It's trying to find a good list of services that use it your major companies seem to do it But it falls off pretty quick once you get outside of your big companies and I see big companies Let's say like google uses it obviously It's easy to get authentication with your large companies. It's all the smaller ones that don't do it I wish they did But uh, that's you know, why I still want to make a video because I I encourage more users to do it those users will also bug their favorite services going Hey, why don't you guys implement fido because from a cost standpoint other than the development time which is non trivial But it's pretty well documented how to do fido and I wish more and more places supported it And even microsoft in a recent write-up they had about being breached They said fido would have solved this problem because someone had breached a 2fa from a text message 2fa they had so Even though I haven't tested every key as long as the keys You have are if you're going to go with the fido fido certified They're good to use and uh, definitely going to increase your security posture And it doesn't have to be ube key and that's one of the reasons I'm holding in my hand something not ube key To show people one there's some other options out there ube key is not bad nothing wrong with them But when you start looking at pricing you go well, I want to buy a couple keys I want to have a backup key great. Maybe you don't want to buy two ube keys I would say for our audience to continue to suggest things that we should look into So if you for example notice another kind of key, let us know But I would also say That you should also let us know what it is about that key that ufa makes it different that makes you curious about it Because it just like this person said this person literally said, you know, you can't check out all of them Which is true But if you help us Understand what makes something stand out that might get our attention better because if it says have you checked out the only key No, I haven't but if the next sentence is because it has this awesome feature that All the other keys they don't have yet I really suggest that you check this out. Okay. Now i'm more curious about that particular key because I know there's something that makes it stand out Yeah, I mean what stood out to me about this key that i'm holding one is part of the phyto alliance two prime shipping It was free I was like, hey, I can just buy this and add the card on amazon. I'm not gonna lie. That is a factor Availability and ease of purchase matters when you want to buy a bunch of these So that's something that falls into the part of the category of research Nothing like impulse buying and retail therapy into impulse buying had a factor in that Yep Let's see what's the next question. Have you thought about a homelab prequel? Not another show but an episode q&a. I want to read this question. I think it's interesting Why you want some vpn? What you can do with it? I think we cover everything from the basics. So we don't really need a prequel. It's kind of a Follow along all of this so Yeah, if we do a prequel that I think we're going to start inviting some of the toxicity that the star wars community has Because they don't really seem to like prequels much and i'm kidding. Of course, that has nothing to do with our podcast but We have specifically Structured the podcast such that someone should be able to listen from the first episode to current And obviously there's going to be some things that are of the times but generally speaking We're using old equipment. So we're not really as vulnerable to You know something being outdated as other audiences because we're using hand-me-down equipment here. So I don't really feel like there's a ton that is really of the times It's just you know, we would we want you to be able to listen from the beginning on up and By that point you should have a decent understanding of what we're talking about by the time you catch up with us in the present Yeah, and there comes a point where we're also not the here's the mouse and here's the power button people Um, I there are plenty of other people out there. It's like that's part of the thing One I prefer to see more towards this is Even this podcast is maybe a little bit lower than some of the more deeper videos that I get into Because we can't convey that as easily in a podcast the same with j j has entire videos where he dives deep into ansible That's also not a beginner's topic So it's trying to figure out where we fit in We did this so we can encourage a lot of people to get in there And we definitely tell people who start here and want to go more advanced But if you want to go less advanced in this there there's plenty of youtube channels out there There's some very 101 great channels, but I don't know which ones are because I haven't watched them I just know people tell me this person helped me get started or this person helped me get started So, um, it exists if you search for things like how to do something really basic There's probably already a youtube video for it. Um, so yeah me and jake start where we start It's always hard when we're thinking about this figure out the best way to do it It really is in my case when it comes to linux. That's you know in the name of my channel So it's pretty obvious. That's what I cover So if I start talking about a networking topic as it pertains to you know to linux There's going to be some requirement of of some understanding Because I'm not for example networking 101 tv If I was to cover generic networking from the very beginning Then that's not specifically linux. I'm outside of my niche at that point And so many other people out there like you said have already done that So I'd just be kind of repeating everything But once you get the basics down Then we'll be right here waiting for you when you get to that point and we'll get you We'll get you all the way up to advanced Yep um Did you want to take a line 96? Let's see disrupting the ad model. So I didn't actually check the ad nauseam plugin as of yet that this individual has mentioned to us um In regards to episode 49 mentioning disrupting the ad model. Um, and they're claiming the ad nauseam plugin Does that I I feel like I guess where this comes from without You know going too far down in the rabbit hole because now i'm just talking about a pure opinion on my part This is not fact I haven't had time to research this yet. It's just a hypothetical thought that I had it's it's like We hold the power ourselves when um when it comes to tracking and privacy because You know the number of people that are concerned about this we can disrupt that by feeding bad information Now don't get me wrong, you know, obviously there's ads on my videos. I you know and paid from these ads You know, but I also go into the ads and I delete a lot of them too that youtube and your injects in there because I found out They're putting ads every four or five minutes And I don't think that's fair, especially when it's you can't even watch the video But a lot of people, you know out there if we do this wrong we have ads all over a blog page You can't even read it. So it's not like we really hate giving money to creators That's what sometimes we could be accused of by using ad blockers But the reality of it is at least on my part If the ad is reasonable, I don't really feel the need to block it if it's flashing on the page It's interrupting my reading which they often do It's a negative experience. Then I I have no guilt in blocking it whatsoever But at the same time it's like if the ad companies are going to drown us Then we have the power to fight back if we all unite but at the same time I don't want us to fight back to the point where no creator makes money at all There has to be a balance So I don't know if that's quite the way to go because I haven't checked out that extension or that GitHub project yet But I think there's got to be something we can do Yeah, and we don't have the same hot take that Linus had tweeted a couple weeks back something along the lines of people who block ads are The same as movie pirates and I'm like, I don't think so. I don't think that's exact That's not a one-to-one equivalency my friend. There's no truth to that at all I mean, it'd be one thing if I was blocking ads just because I can Ads never bother me until they do if there's a banner up there or you know something before a video plays I don't mind that so much because I get it because my business is built on that But here's the thing if you go to learn linux.tv and you see ads all over the place You can't see the video. You can't read anything and you have to click away like 10 different things You know what blocked my ads. I deserve to lose money period and a discussion. That's doing bad business And it's creating a bad user experience So for someone to say that I feel like they're generalizing things because yeah, if we blocked ads just to block ads That's bad But what we're experiencing here as users and it's important for content creators to understand this Is that we're reading a blog article in our free time But it's hard to do that when you have things popping over it constantly So all we really want is a balance between You know content creators making money But also having a good experience on our end and that's why we do it I think linux should probably understand that it's not like we want to steal his content We just don't want to have 10 things over top of the page that we have to remove Every time we want to read something. Yeah, it's finding a happy balance on there And it's not easy and the ad market has always been terrible and it's not better now It's just less pop-ups now because browsers the day that they decided not to have pop-ups was great But here we are today with just a different problem. So we have pop over at we have popovers now So we don't have pop-ups. We don't have pop-ups now. We've got pop-overs. We've come full circle We're getting really bad. When's the last time you've read an article online and you know one minute in you didn't have something Pop over right in front of what you're reading. I can't remember the last time I've done that or we're able to do that Definitely go back to buying books That's how you can help authors and creators and read a really solid good book We had a good discussion over the weekend with our friend Michael Lucas Um discuss maybe we should Michael Lucas on here talking about learning and technology and things like that We certainly recommended his books enough. We'll we'll uh take that into feedbacks um, yeah, all right next question not related to advertising is Would love would love to see one of you guys talk about installing zero tier on true nasty scale well I uh, I think I can feel safe in in uh quoting what wendell told me when we were discussing some of the q nas scale problems true nas scale problems, um is It's still tricky I think the words he used if you color outside the lines a little on docker and things like that How it still breaks things. I don't know how well true nas, uh would integrate with like things like zero tier It becomes kind of tricky and to me It's not the place to really run it But I guess if you are trying to build your network on zero tier. I'm still big fan of vpn and I vpn things to Uh, when I'm outside the office when I'm inside, you know, and I want to get outside I vpn to the things I need and that's how I do the transport layer putting things like zero tier directly on your nas If it's not built to handle it and the People building it ix systems didn't think to integrate it. It can be a lot trickier to integrate and then also break things I've already had some problems. Um, this is why I've been slow on the true nas scale videos I have questions that I can't find answers for in documentation. I'm a Rtfm kind of person So when I keep finding some of these problems now it gets even harder for me to integrate because we even found me and I was hanging out with wendell from level and text for the weekend along with j and this is all a discussion We had about some of the challenges with this new product and Not being able to get certain network aspects features to work not even related to Adding something like zero tier to it. So I don't know how soon we would do a video Relating to that and also if the until it's something that the ix systems people want to start baking in It's going to be a problematic thing to add it There's probably going to be someone right do a good write up that's taken the time to really dive deep into How to start modifying it the problem is with any of these when people start modifying a core function of it to Bolt on some functionality the update monster comes and breaks all your functionality and you have to redo it again with the updates And that's less sustainable to me versus Whenever there's no update to true nas I just update it and don't think twice about it because i'm using a vpn again So it's unrelated to true nas now the true nas does have open vpn And I think wire guard natively built in so you could possibly do it that way for other ways to think about connectivity If there's a reason you want to connect directly to the nas But back to from a design standpoint generally you want your vpn because it's usually not one thing I want to get to on my network i vpn to my network I get to all the things on my network instead of having to do each device Individually so I have a synology on my network. I have a true nas on my network I have a home assistant on my network and when I vpn i'm accessing all those things As opposed to setting up a vpn to each of those things granted zero tears another way around that But I don't know if it's the best way so there's no So I feel like this is a very common element of the life cycle Or the you know process where do you want to call it how a homelab person evolves over time So if you're a beginner and by the way, I'm not alleging that the person who wrote in as a beginner at all That's not what I'm saying, but in a general sense Speaking about nobody in particular You might be and I was you know personally of the mindset. Oh, I want to implement this thing I'm going to add it to all the things so all the things have the thing And then as you get more experienced like you were basically saying Is that oh, well, I don't have to have it on all the things because if I put it on the router or the firewall Then it's basically On all the things and routing will get will make things go where it needs to go So you could put put it in one place rather than a bunch of different places Which makes sense. It's easier to maintain And that's something you generally want in a firewall or router Anyway, but the other opinion that I have on this which could be a little controversial I think a lot of people will agree with me, but some won't and that's totally fine When you decide to Set up an appliance by appliance. I mean a turnkey solution Obviously true nas has a bunch of dials and things that you can set But I consider true nas an appliance which means you're using something as Created by the developer for the intended purpose of that thing So if you're going to go in there and start adding this adding that Customizing this then you're basically in my opinion making an appliance into a custom build and if you're going to Do that why not just jump straight to custom building something because that would be what you want to do anyway Now I used to be you know earlier on adding ansible to everything even like proxmox for example But I stopped doing that because proxmox is an appliance I'm going to use it as they intended it to use Or be used and if I decide that It doesn't work for me. I need more features or I need more control Then I'm just going to go straight kvm q emu on a custom built server and just roll it all myself when it gets to that point And the other thing to consider on this too One of the things that's important to understand is like there are certainly projects that say start with a bunch to start with Debian load our thing on top of it. You can even look at something really broad like WordPress I'll tell you here's a prerequisites for your base operating system when you look at something like proxmox or true nas Whether it's scale or core They are Fully rolled in you don't load the os and then load their tooling on top. They're fully complete. They're loading the tooling They've done so much customization They don't want you starting with the base os and loading Their libraries on top of it and their scripts and all of their code on top of it They want you to roll it as an installer and the reason for that when you start digging into the way they did Debian They stripped out lots of features To minimize it and they also customize lots of things in the way it works So you won't find things in the same expected places all the time Which creates these challenges when you're trying to integrate because normal tools expect things to be based on normal Debian and true nas scale is not normal Debian. It's Debian customized to the way true nas scale works So that's an important when you're thinking about any of those That bolt-on functionality. It's just it's not always impossible But it's always off to us should be often I should say a clue of trying to get things to work together Yeah, I have you ever noticed that the mindset generally seems to be when it comes to a product that uses Something like Debian or Ubuntu and underneath That you might see comments like well, this is just a sent to us Reskin or this is just a Debian reskin and it seems like it's everything like you use popo s So that's just an ubuntu reskin. You buy a system seven 76 computer That's just a clevo even though they do considerable work to differentiate that and popo s is super different than Ubuntu now just because it's built on Debian doesn't mean every Debian design Mentality was followed just like you were saying and then you could create a situation where you're in the forums You're trying to figure out a problem just to find out that your custom chef implementation is the reason why it doesn't work And now you tied up the developers and community members to try to figure out A problem that technically shouldn't have even happened in the first place Yes, and the rule of thumb is really simple for any project so we can speak more broadly to this Does the project require you load the base os or does the project come with with the base os it comes to the base os It's probably not meant Oh, it's not I would say probably going to be harder to mess with and harder to implement customization to it If they have you load a base os probably going to be really easy to customize because you load the base os And you can probably throw a few extra things in there to make yourself happy So that hopefully clears all that up for all the other projects because there's so many from out there That this question could be applied to Yep Next one is true nas scale. I'm sorry not to scale tail scale the other scale thing So we did a zero tier. I don't know a problem. I can do a revised one on tail scale I've done a video on it before we can throw it in our queue for once the cover tail scale is interesting It's a management plane to control wire guard as the fundamental back end to tie things together works extremely similar to the way that Um zero tier works. It's a mesh network if you will that has some really clever features It's really popular. It is also a commercial solution. So so a zero tier to an extent Zero tier seems to be more open source friendly tail scales open source on the client side But very clearly proprietary on the server side. So maybe we'll throw that in the queue to eventually cover it Yep Yep, absolutely. All right. I'm gonna skip the long list. Someone just had a long list of suggestions for things to cover Um everything from docker to rsync to bit rot to that. So um, yeah, those are good topics Yeah, all good topics on there. I don't know if we're gonna cover open wrt Um, so they threw it in there and I don't really I don't know pf sense is so inexpensive It's kind of the go-to in the home lab because of the flexibility and the fact that you can buy Their hardware device is relatively cheap or you can just load it on any older hardware you have So I don't see the open wrt I don't use it either. It also would be harder for me or j we me and j talk about this like offline a lot Sometimes it's hard for us to do a product that we never use or don't have an interest in using It's hard to do the video because we're not going to be experts in it because we don't use it. So Yeah, but the other topics such as um, self science certificates, I You know after attending the talk recently on michael lucas over the weekend on ssl and tls and certificates I think maybe an episode just on how that works is Definitely worth diving into uh as a talking point. So I literally bought the hard cover of his book while I was there So yeah, yeah, it's it's definitely one that I think there's even I'm working on a video about how ssl inspection works because it's a point of much confusion to people of actually how Uh as tls inspection works How man in the middle works how you manage that in enterprise environments and also how you Deal with things like s and i and what happens when we go to es and i what breaks next and how you break tls 1.3 And why sisco wrote an article on why they hate tls 1.3? Yes, sisco that sisco did There's a lot of drama there. Um, there is I was gonna say i'll to pick you back up what you said I mean, we yeah, we would love to try all of these things and maybe a number of them We will sometimes things can be in my queue for several months before I get to them But it's also possible when I find an editor for the videos that I do on the youtube channel and that time comes back to me I might then be able to dive into those smaller topics and Create more content around those things. I mean, I'll get to the content regardless eventually because I have a queue for a reason but Like you were saying if it's something that I use already It's not that well everyone should be using that thing. So why would I look at anything else? It's like low hanging fruit. I already know the thing So it's easy for me to do a tutorial on it because I know it But at some point I'm hoping to free up some more time and maybe I will Explore outside my comfort zone more often Yeah, yep. It takes time to get all these put together I'll take the next one, which is very simple We were asked about I don't know if it was targeted towards you me or both About open stack. So the short answer is most likely. I don't want to promise it I'm not going to give anything away right now But I am actually talking to a company that wants to give me access remotely to an open stack stack And I'm going to be meeting with them this week to see if that's something that they can do And if it is then I'm going to consider doing some videos on open stack Actually, that'll make it a lot easier for me to do that considering the hardware that you might need to run it But I'm not going to promise it because there's all kinds of video topics that I wanted to come out with and didn't work out But my intention as of recording time right now is I definitely want to do that I am taking action on it and engaging with that company and if it works out I'll tell you what that company actually is and then give you guys like some kind of an eta I'm when video content might be available But I've always liked open stack. It's just been one of my my geeky things that I like, you know We always have our thing like why do we like that? I don't know. I like it. I like it because I like it and I So I'm hoping that it works out I will hope to let you guys know Just keep in mind, you know, my my backlog is usually three months deep So it could be summer or fall before we see content around it But it is my intention to make open stack content. So we'll see how it goes. Yeah It's definitely pretty cool. Jay's the expert on that more so than myself um I'm going to mention this one that we're not likely to cover it with with me or j which is sef fs sef and gluster But me and j are not like to cover it by ourselves. That's the key word I do have people I know I've talking to the 45 drives people About doing a video. I don't know if it'll make it for the podcast It may be just a separate video on the channel Because because of the complexities it needs some I don't know how good it would be to verbally explain We'll we'll chew that one a little bit to see if it's a podcast episode But the other problem is and part of what we're going to discuss And this is what me and the team at 45 drives came up with is not everything needs to have a clustered file system Despite the number of people asking Um, it's not the solution for everything and that's something people don't seem to understand So we we'll start it with the perspective of do you really need a clustered file system to do it as a learning process? Absolutely jump into it. Have fun. Um run into those brick walls of oh, wow This is a little bit more complicated But because it's an extra layer on top of things you don't format the hard drive sef you Build your storage servers and you load sef and you set the storage nodes up And there's all the things you do but all this extra complexity It does have an absolute use for things like, you know enterprise high availability extremely large storage servers But I get people suggesting that I do it for even like when I I did a video Talking about just setting up one server for a client. I'm like, yeah, that's not their use case They don't want the complexities that come with adding sef sef on top of what they already do So it's not just because they have a petabyte of storage doesn't automatically mean but they need sef That's the only way to do this no because now you have extra complexities that would have increased the job cost I mean sure if we have an unlimited budget just buy three or four of those petabyte servers and I'll throw sef on them and cluster them together. Oh build all the redundancy and the networking They need for that which they don't really need but you know the budget's unlimited in people's head So it makes sense to go ahead and do with it. Oh, and then just bring on assisted men who know sef to Make sure everything's right. So updates go fine and all the stack. It's updated properly Because all that complexity, you know, we got a budget for that too to Have maintainers for it or just selling one sure next server with uh a petabyte of storage because If the storage even was down for three days, it would just be annoying but not put the company in a standstill So you got to think about when those use cases are and that's kind of how we're going to start um That some of that talk on there it's not for everyone don't try to design an overly complicated solution That's hard to support and of course hard to support also means has some expenses that come with it When you do things in the enterprise world budgetary concerns because you you want to unless you're a bad at designing enterprise stuff You don't just design it you design it and you plan to support it There are plenty of people out there that Got the hammer and nails and propped up something really rickety and that's what they call me and jade are like Yeah, the person that set this up, uh, they vanished. Can you look at this and me and jade was like, um We need to start because they did what with what um, but I think my one of my um One of the annoying things in the industry Industry wide is the mentality that you learn how to use a hammer now everything's a nail So you have to use a hammer and everything um Now I can confirm that i'm including myself in this maybe not nowadays because i'm not like that now but when I first started Um, you know, for example, I learned python the first time quite a while ago I had to use python on everything including an rsync script that I wrote which was really good if I do say so myself But it also has probably about 50 lines more code than it would have if I had written it in bash Which is more appropriate for that kind of thing. You just got to use a solution For its intended purpose and don't be all in to the point where you're forcing it to work Um where it doesn't I used to be that way like I said, you know for quite a while I mean I would literally when I was You know very young and I wanted to I'm building a project and I don't have a screwdriver But I do have a hammer and have a screw but I can't find a nail and I can confirm with enough effort And you know being completely focused on it You could absolutely drive that nail or that screw into a board with a hammer But that's not the right way to do it. Um, and it could be I was just too lazy to go upstairs and find the screwdriver But my point is that when it comes to it it's like we need to containerize all the things everything must be docker We can't have a single virtual machine. No, no, no guys. No, there's a use case for a virtual machine There's a use case for a container. There's a use case for sef They have their place where they fit and if you force something to work where it doesn't it might work great at first But eventually you're going to run into problems. Yeah absolutely, um nested kvm virtualization I don't know why they put nested in there, but uh, definitely some more talks on virtualization Yeah, I'm not nested virtualization is where you take Uh, let's say I'm using xcp and g and I load a bunch to inside of it And then I run another virtualization inside of it. Yes, that's supported. That is something you can do with xcp and g I think proxmox supports nested virtualization hyper v and vmware do as well But I don't know the person's asking really about nested, but other virtualizations I've been using this and someone asked about it. I gotta admit I didn't know this until more recently because Ain't broke. Don't fix it. I use virtual box on my desktop to Run things and it runs really fast It did a video yesterday on true nascale permissions and you know rebooting my windows 10 machine in my virtual box on my linux Takes about 15 seconds. So it's certainly to me fast enough But I do know the menus have gotten better for some of the other virtualization Like kvm. So it's worth maybe someone will talk about as virtualization on your desktop It's not a topic we've covered yet where You have a desktop you go. Well, I don't really have time to buy another server But I want to load something one of my go-tos has been virtual box because it's free It's easy with their menus out there. It's not hard to understand And it works cross-platform. So when I talk about virtual box, you could be running windows You could be running linux and it's the same virtual box But there are other ones out there and they've now got a better menuing system for some of the other ones I've learned so uh, yeah, that might be another topic we cover because I think that's another handy thing That you have a good desktop Maybe it's your gaming desktop and when you're not playing games You're not using that 16 gigs of ram you put in there You're not using that fast processor all the time and you like to do some learning on it And buying a server really isn't fit in your budget or fits in your dorm room Wherever you may be you may be of limited space and as much as we encourage you to go buy something on linux You're like, well that also has a cost to it as well. You already have this device I still boot up linux VMs because of the convenience of having them in virtual box It works really well But that topic in general of running some other type of virtualization system on your own computer simultaneously It does work really well and um, so something we can definitely explore in the future All right, uh, let's see The other I think the last question we have in here, um Oh, is chunas scale reliable enough is the last question we have in here Um, I haven't found with chunas scale any reliability problems Most of the problems I've run into are speed related. We it doesn't seem to be as performant as, um Chunas core is I think it just hasn't gotten optimized yet Reliability wise provided you use it as they intended Note my earlier comment of if you color outside the lines a little as windows said Sometimes you'll end up breaking things and docker And breaking all the things that you have running on it So if you're using it as they intended and not loading third party Modifications to it. It seems to work quite well. Uh, we've had a server running for a little while We've been doing testing on and functionally it works Um, so I don't think there's any reliability in it But if you're looking to squeeze the most performance out of the hardware you have That has been the drawback as of right now today Still april of 2022 future versions are Undoubtedly going to go through just like chunas core has optimization over time. Uh, we're just not there yet They're they just got released a month or two ago if even that so I I have figured based on the conversations that we've had on camera as well as off camera And we've talked about it this weekend as well. I've talked about a lot this weekend. We were hanging out all weekend. So um I just figure maybe I'll give true nas scale another look at like version three or something because right now I mean I was planning on switching to it because My mentality was you had me at linux, right? Um, but When it was slower when it was released than it was either the beta of the release candidate Okay, I can't really get behind that right now. I need to know that the technology Is more solid, especially if I'm going to be relying on it for the back end storage of my video editing Routine that's not something I can really mess with because that will literally cause a huge problem So but I do want to take a look at it. It is linux-based. So I absolutely will try it out again But I just want it. I want it to have more time in the oven Yeah, the um someone mentioned too I'll go back to the previous question Wendell has a probably a good write-up on this where you can turn your system in so it boots off of an os such as Xcp ng or one of the other Hypervisors and then starts up another operating system that projects to one of the screens with multiple cards So you have multiple video cards in it. There are ways to do it. I think that's what they're meaning by nested That's what I'm assuming based on the comments. I see coming in the live stream here. So um, just to go back to that Yeah, I I don't think we're going to cover that as a topic. That's a that's a very complicated and fun project Way harder to do in verbal words Um inside of a podcast that's a that's an explainer and a write-up of how all that works I think it's even difficult to do in a video because if you're nesting something with something else Then what are the odds that someone's going to be nesting the exact same thing? And you're you're doing in the video because the whole definition I mean you're nesting things It's because you have unlike things or maybe they are like things But you're combining things in a way that may or may not be how it was intended to be used And what it comes to tutorials? We live or die by how reproducible That is when you watch it if you watch it and you get a completely different experience Either you've done something wrong on your end and you weren't following it properly or I've done something wrong on my end And I didn't explain it properly or something who knows right but in that situation Tutorials are tested one to one this version that version But it gets a lot harder when you're nesting things in my opinion That's kind of hard to not impossible, but difficult to translate into a video Yeah, it's it's a pretty neat. Um, if you Go over to the level one tech forums There's some good write-ups in there on that whole subject Of doing it because a lot of it has to do with people who want to even do things like video game passers So they want to run linux But still run windows and pass through windows boot from it and then pass the video card through in a meaningful way So it doesn't lose performance so you can now play and use your windows nested But you know I mentioned virtual boxes the not the best way to do it because if you don't get the pass through the graphics card properly You don't get the performance and you know gaming is all about using that graphic cards you have for performance So yeah, there's some good write-ups on there. Um Yeah, it's a great topic, but it's also Uh harder to cover in a podcast So and you know the thing is too, I mean I will say there there is some fun Aspects of this where I can't remember the youtuber It wasn't that long ago Where they literally had raspberry pies running in a desktop case along with an x86 board desktop computer board and and They were part of the equation in the same, you know Frankenstein Case of mismatch, you know chipsets and parts. It was so freaking cool. I mean it was so awesome I don't know how reproducible that would be on my end, but that's some great content But outside of that there's really only so much we can do and but anyway If you out there have some really amazing combinations that Are just really weird strange cool or epic. Let us know please if you really want to know about it Yeah, whatever you created Yeah, feel free to fill out the contact form on all of that and uh We'll go a few more minutes and see if there's any questions here that we missed in a live stream chat Do you see anything in here j here? Well, let's look so I think we got that in of it We did the sef questions People talking about virtualizing their firewall, you know from a homelab perspective I like virtualized firewalls from a business perspective like enterprise-wise Keeping hardware hardware's in it comparatively speaking in enterprise spaces less expensive So yeah, I still I'm still a big fan of not trying to virtualize all your firewalls Yeah, I used to do that. I think at the very beginning and It's not impossible. You can do it It's just you'll find out some edge cases pretty quick that are hard to explain that you can only Experience by trying it and it's not that you can't work around those things because a lot of people do and they have no problem working around it But it is going to cause you to think in a way that's going to be oh, yeah, I didn't think of that So um, especially when you have a single point of failure and then everything comes crashing down That's not fun either So I'm a fan of keeping something outside of the or the firewall outside of the virtualization stack But then again, that's just my opinion And if you are able to do it and you maintain it and it works for you then no judgment. Yep Uh, there's people that did ask the question I see someone said I want to know how much of a difference there is between the new ubuntu and the old one I'm assuming you mean the last release good news our last episode is for you We covered the differences and changes in there Um, and on top of that jay has reviews. He's released on ubuntu with some spicy opinions Yeah, but I mean I think in terms of ubuntu server, which I think is going to be the most relevant to our audience here I mean we talked about the desktop version Pretty much solely because there wasn't much to talk about in regards to the server version Which you can argue when it comes to a server operating system. That's a good thing There's there's just not a lot to talk about great. It works The desktop version is an abomination that I will get into I'm very disappointed in canonical and the release of ubuntu in the direction they're going But I'm not going to say any more than that But when it comes to ubuntu server, it's still solid It's not that different It compared to 2004 if you're coming from an LTS you have a newer kernel Actually newer pretty much everything the technologies have been moved forward but I don't feel like it's a huge change to the point where I'm creating or actually updating the mastering ubuntu server book which is happening We are right. I'm writing it and we're going back and forth with the publisher right now I don't have a release date But what I can tell you is there's going to be less of a change in this edition From the last edition compared to, you know, the second edition to the third edition There's going to be fewer changes. There's still new content But it's not as big of a difference because with 2004 there was a lot different. We even had a new installer I had to rewrite the installation section While we were going through the process because ubuntu changed it at the last minute But now everything's pretty much stable and Solid so I have no hesitation recommending ubuntu server at all I still think it's one of the best along with debian for servers. Yep I see Veronica explains this in here if you don't follow her channel Check her channel out as well and Veronica says I don't virtualize client firewalls in production Too many bad experience with failure points on the host server We actually had a really weird one. It was really annoying Someone with a virtualized firewall you couldn't ping some devices and some devices couldn't talk to the firewall We don't know why we do know what the solution was. It wasn't restarting the firewall service Uh or the vm the firewall is running on that didn't solve it It was though we still we can't reproduce the problem the way we solved it was shut down all machines on the vm Reboot the entire vm system bring it back up And magic all the things that wouldn't ping ping We do know looking at the arp tables that there was some weird assumption the arp table was somewhere else We don't know why we kept getting a bad response We maybe could have restarted just the network stack I don't know the best thing we did was just reboot the whole host vm to solve it These are rare occurrences But what you've done when you virtualize the firewall you've added one more layer of complexity Um and in the enterprise environment you want the layers of complexity to be minimum because the goal is to maximum uptime maximum support And putting in ha firewalls sometimes even to not necessarily because you are worried about the firewall failing But also so you can manage updates and have rollbacks are good ways to do it. So You find out some like I was saying earlier some weird edge cases one of which might be Yeah, I need to back up my pf sense vm. So i'm going to just shut it down I'm going to take a backup image of it And make sure that I have a way to revert back because you know having backups of the config isn't enough You want a bigger backup than that, right? So then you're like wait, why can I back this up? It isn't working. He's timing out. Oh, that's right My storage server is somewhere else and proc or pf sense is avm But since I shut down pf sense to back it up It doesn't know how to route to the storage server anymore So how do I get the backup from the local storage to the backup server? What I could do is of course just save the backup locally then bring You know pf sense back up and then I can route again So now I'm able to get the backup where it needs to go But you'll find all kinds of things like that that you're not always going to think about when you design this that In my opinion are just headaches that I would avoid If you can to the point where if you could find like a you know an old desktop that has a You know two one gig ethernet nix in it or um, just just put that in there It's gonna save you yeah, it might run you a hundred dollars used if you don't have one lying around But it's still a lot better than spending hours trying to figure out why you can't do something just to find out that The design it doesn't really scale to how things are normally done. So you've got to run into edge cases It's frustration and how much is your frustration worth? I think it's probably worth checking the local yard sales, especially as the weather gets warmer in the united states Find something to install pf sense on and put it in put it in your network You'll you'll you'll thank me for it. Trust me. Yep Jay, have you got your released video yet for the new version of papa west or you working on that one still That one i'm hoping to put out today It's 90 done, but I decided to add another scene at the last minute, which Um, I was about to hit the render button right before this started, but I was just a minute too late I'm gonna try to get that out today Yep, perfect because I seen someone asked about it. Um, so that video is coming and we'll leave you with that That's uh, yeah, check that out. Keep keep an eye on jay channel. Make sure you're liked and subscribed over there Uh, so you can know when that comes out. It should be out Like jason within the next 24 to 48 hours. I know how he is. He's obsessive and won't stop until it's done Right and it will be today as long as that random bug in youtube doesn't happen where it's processing for like two days straight Oh, I know because it you know, I could upload it right now Which which mic with my connection will take an hour no problem But then after it gets uploaded it might be processing processing and then days later Like are you ever going to finish? But most of the time it's within 15 minutes after that But I you know, it's up to youtube at that point how long it's going to be processed So I guess we'll find out Yeah, that that is a challenge we creators run into sometimes there's this little spinning wheel That says processing and then it says five minutes left and then it says Taking longer than expected and there's no eta. It just that's the notice we get because like we'll upload something You'll have to wait forever So I literally wake up the next morning and and see it's still the same status and the next morning again I'm like, oh my god, and I look on twitter and everyone else is complaining too. So I know it's not just me So yeah, hopefully it'll be soon. We'll see how it goes Yep. All right. So thanks everyone for joining us and uh, we'll see you next time