Transcript
the number of the Veeam Community Recap. I'm Rick Vanover, the Rickertron, and guess what? I'm in normal circumstances. Maddy's in normal circumstances. Maddy, how are you doing today? Hello, Rick. Hello, community. Yeah, I'm doing great. Better than last time. It seems like my throat is still a little bit sore, but much better than last time. But I'm doing good. We are very busy. We have a busy day ahead, full of good stuff, I would say, and hopefully we're going to learn something in the process. I'm not going to say more than that, but yeah, I'm excited about it. Yeah, indeed. Things are running very busy around here and around there being our community. So I think I've got 12, 15 tabs. I don't even know. So I think we should just jump right in. You ready? Yeah, let's do it. All right. Here we go. I'm still figuring out this new tool. I'm not the quick switcher like I used to be anymore, friends, but we will start right here with our first content from Antonio D'Andrea. Oh my goodness. Did I get it right? Yay! Yay! All right. Antonio was talking about the Veeam Agent for Linux and a snapshot kernel error that is happening for him on Oracle Linux 8.9. I got a few thoughts on this one, but Maddy, what caught your eye here? Well, apparently this is a classic Linux Veeam integration issue as far as I researched and I found out. So Antonio is trying to run backups with Veeam Agent for Linux in managed mode with the application aware processing enabled. That means snapshots are required, right? So the job keeps failing with those errors around fail to create volume snapshot, kernel module not found, and so on and so forth. And at first it looks like, you know, it could be a typical compatibility issue between the kernel and the Veeam snapshot module. Even the official KB, and you have a lot to talk about, but that's another KB I think, points in that direction. So what he's doing, he is kind of following the expected fix, rebuilding the Veeam snap module with a newer GCC version, but then the compilation keeps failing because the system keeps failing back to an older compiler. And as you can see, everything pointed to the kernel compatibility or the GCC mismatch, which are okay valid assumptions, but the actual root cause was actually a layer below that. How DKMS handles the build environment. So I think we have to be careful, because in this type of situation, sometimes the issue isn't the thing that's failing, it's the tool that's supposed to fix it. So that's very interesting and I want to thank Antonio that he shared it with us. Now, all your thoughts. Yeah, my thoughts are like, this is an area where the updates make the difference, because the tooling, it's the nature of Linux, where the compilers and the different modules have assumptions of the environment around them. And by no means am I an expert in Oracle Linux, much less 8.9, but I do know that one of the first and last things I always do in my Linux operations, I do a lot with Ubuntu, is just do the update and then the upgrades, right? Get app, get update, and then get upgrade. Either automate that or do it regularly, because those are the things as things go out of over time. And who's to say that a system didn't start with Veeam installed on it, right? Just those components, those modules that in this example, the compiler and the platform it used, right? So it's really two things, right? So between GCC and DKMS. So not a Linux expert, but I know it's always good to update. Antonio D'Andrea, thank you for sharing this specific scenario. Totally. All right, here's a good one. Esteban is talking about Red Hat Summit, great event recently in Atlanta, and with Veeam to protect the AI system, fancy in production. What caught your eye on this one? Oh, great to mention Esteban, first of all, from Argentina, one of our legends, back on the community hub with more great content. And this one actually sounds a little bit familiar to me as a concept more than anything, because I wrote a blog on digital sovereignty like two weeks ago or last week, and I posted at the hub last week. And Esteban, as you said, he was at the Red Hat Summit, and he writes on how we should think about protecting AI workloads on open shift. And the main idea is a shift in thinking, what he says, AI placement isn't just a technical decision anymore. It's about control, risk, and where your workload should actually be. So he introduces this concept of sovereign AI, which kind of goes beyond just data location. This is a point that I made as well in my blog. So he says it's really about three things in here, how to control the platform and the models, controlling the data, and the most important to be able to actually operate, audit, and recover everything when you need it. So the key message is that AI workloads aren't just containers or storage, right? They're entire systems made up of, you know, pipelines and models and metadata, secrets, vector databases, Kubernetes objects. And if you only backup the storage, you're not actually protecting the application. This is something that he mentions. And the example he uses is fine tuning a llama model on open shift shows how easy it is to miss critical pieces like pipelines, CRDs, secrets. And that's where he talks about Kasten. He says this is where Kasten comes in. It treats everything in the namespace as a single application so you can recover the full environment, not just parts of it. He says, you know, you are used to think backup the data and you're covered. But obviously with AI workloads, and I think we kind of learned that as well when we talk security AI, that's just not enough anymore. If you restore only the storage, not the pipeline or the configuration or the model metadata, the rest of it, it won't actually work. I think I really like that sovereignty angle. I think it's very interesting. It's not just about compliance, but it's just about control. And there are some really good questions, you know, to be discussed in there. So I really like this one. What did you think about it? I, of course, like it as well. And I like this visual here. I think this kind of shows the really the way this can live, right? And the thought here is that the kind of the main workload we care about is the open shift container platform, like in the middle, and then protecting it with the Kasten. And what I love about these, you know, resource definitions and these AI workloads in general is they're going to be a collection of a number of different things. And whether it's, you know, a physical server, a VM, or a containerized application powering AI, the moment the business depends on it, it has to be protected. So I think that this is fantastic. And I'm going to go off script and, you know, like, for example, vector databases with we did this at KubeCon one year, you know, if you also want to watch a video, it was on YouTube. Oh, goodness. Anyways, that's what I get for trying to find a video I watched a year ago. Okay, that doesn't end well. But maybe I'll put it in the notes. But we have a YouTube video that Topeka and some of our dev team have done on how to run vector database backups and more. So there's, you know, a lot you can get into. But this is really a good start here. Thank you, Esteban. You know, really deep stuff, right? The thought is, once a company builds an AI solution, it needs to be protected, and not just a container and a VM, but the whole data as well. So complete story. Good stuff. All right. Next up, Christian, I am so ever so talking about using blob storage as an object storage repository in Veeam with Azure. Love this one. Yeah, absolutely. It's great. But first of all, I want to give a shout out to Chris, because he's killing it, you know, in that Oxford style debate. He's just bringing it forward. We have like over 70, you know, comments in there, which is great. So shout out, big shout out. Thank you, Chris, for all the input. But yeah, this is a great piece of content as well. In this one, he talks about how Azure gives you a scalable cloud native place to store backups and how it starts with the basics. You know what an Azure storage account actually is? Basically, he explains it's essentially the container for all your services, for all your storage services in Azure, like the blob storage for objects, file shares, queues, tables. And for Veeam specifically, the relevant part is, you know, the blob storage, because that's what acts as object storage. And the key idea is that instead of storing backups locally, or on traditional storage, you can use Azure blob as a cloud based repository. So that's the point that he's making. Through this, he has a link where I think it takes us to the main blog where he's explaining everything in more details. But yeah, Veeam supports this model pretty broadly. And you can use object storage as a backup target, especially for the long term retention, or as part of a scale out repository. So yeah, he's also touching, if you look at that table, he's also going over storage account types, standard premium tiers as well. So yeah, it's short, but actually it contains a lot of good information in there. And if you want to read more, just go to his blog, to his website, so you can get the full picture of it. But great stuff, as always, Chris. Thank you. Yeah, indeed. Thank you, Chris. Makes me think, okay, I'm going to find something this time that I know I can find. So when you go and make Azure storage accounts, you want to be real intentional about the names. So I wrote a blog a few years ago that I'm going to pull up here. And this article from Chris made me think about this, but basically Azure storage account names, just a general piece of advice, come on internet. Azure storage accounts, the actual accounts, have to be unique within Azure, right? So they, number one, you can't do something like test or cloud, that won't work. Even though it's within your own account, trust me, this is my blog, but I thought it would work. But there it is. Okay. So I made this, actually, it's an Excel sheet. So if you ever want all the characters here, you can actually make a self-describing name standard, right? And you can modify, you can download this document here that when you create your storage account, it has to be unique within Azure. And so I would just recommend you make it self-describing, and you can even do things like if it's immutable or not or whatever, but I'll put a link to this in there too. But thank you, Chris. And sorry to go sideways on the content, but at least the internet didn't let me down this time. Yeah, at least that. All right. So up there with, among the coolest Vanguard blog spotlight names is Thomas Magda from Poland, anti.expert. Absolutely love that. This is our Vanguard blog spotlight. Yeah. I mean, welcome to Vanguard blog spotlight, Tomas. This is the first time we are mentioning your name, but I just wanted to say that I've been working with Tomas a bit more because he's not only a Vanguard, he's also a VUG Poland leader together with Pavel, the other leader. So if you are interested to hear more from him on top of his blog, they can have the first virtual VUG Poland event on June 11th. All the details are on the events section. So I had to mention that. Sorry. Now back to this, his blog, which is amazing. And I think this is a quote that I'm going to and I think this is a question that everyone at some point probably asked, you know, which, you know, why is Restore so much slower than we expected? I'm sure you asked yourself that, Rick, right? Yes, but keep going. Okay, keep going. Then you can tell your opinion at the end. So the key idea of, you know, of this is that Restore isn't a single action, right? It's kind of a full process. And there are multiple steps like reading metadata, rebuilding the restore point from full plus incrementals, transferring through proxies, so on and so forth. You know, there is a listing here. And the important part is that, you know, your restore speed is only as fast as the slowest step in that chain. And I want to see if you agree with that later. Keep that in mind. So Tomas makes a really interesting point about where actually the bottlenecks are, because most people would assume it's the network. But what he says here, it's like, in reality, the most common bottleneck are actually the repository, the target storage, sometimes the proxy. And only after that is the network. And he says, especially if that is local. So even if you have a fast network, you can still have slow restores if your repository can keep up. So that's very interesting. And one last important POV here, backups are easy to test, but restores are not. And that's exactly why sometimes they get ignored. What do you think about that? Well, I think it's good to kind of explain these things. And I really like this kind of line here around, you know, the repository being part of the scenario. And I like to say that these three things here are kind of what I call reconstruction, right? It's all the work to reconstruct the task that needs to happen. And I remember the 2021 Beat the Ghost Dev challenge, and they were really finding the absolute fastest way to do it. And I think when it came to restore, for example, there was no backup. I'm sorry. I believe there was no compression going on. So like one of these steps were just skipped, right? So it's one of those things that you can do to kind of get around the performance. All of these components are part of the flow. And I've even requested, I have an internal feature request, the bottleneck analysis that we do on backup jobs. I'd love to see it on to kind of help, yeah, help organizations get a good sense of that flow as well. So no, this is good. And I think the only way to find out is to test it. And I think it was Michael Paul one year that did like some AB testing of the different configurations and then how it did on the overall speed. But the big, unique thing is the source data, right? That's the hardest thing to synthesize. It's pretty easy to configure the Veeam settings, but like, what are we actually backing up and how does it move? How big is it? How much does it change? Those types of things. So what I think Tomas really gets in here, this is one of my most important parts here. Don't have it as a surprise on restore. Have a good sense of how your restore performance will go and what tweaks you can make ahead of time. Anti-technical, business and practical. Now I translated this from Polish, but translated pretty spot on. So this is good stuff. Thank you. Yeah, totally. Thank you. And welcome. Hopefully we're going to call your name. I really appreciate his work. Really good. Indeed, indeed. And okay. So now we're going to go to special departments. Still got many, many, tabs here. So let's go through it. And the first up, VUG Brazil, early adopters. They're going to be going on about Veeam and security AI. Looks like Joaquim, Philippe and Ricardo. Yeah. I mean, what can I say? Our community keeps busy and comes up with so many great things. So I'm so grateful for that. Love it. Look at that number, 46 already registered. This event is June 3rd. So you still have time to register for it. I think it's going to be great. It's going to be in Portuguese. So keep that in mind. Now you translate it, but it's going to be Veeam and security AI. That's going to be so interesting. I don't think we touched that much yet of the community hub. A lot of, we didn't have a lot of events on security AI, or we didn't have that many blogs written. So I think this is kind of one of the first, I would say. So great stuff. Thank you, Joaquim, for putting this together. Really looking forward to it. And I'm sure for the number of people that register, I'm sure everyone is looking forward to it. So good stuff. Yeah. And this should be the first of many really of the combined portfolio. So good stuff. Thank you, Joaquim. Thank you, Philippe. And thank you, Ricardo. This will be great. And yes, it will be in Portuguese. I've been using the translate browser because this one will be in Dutch. And that's the Veeam user group Netherlands meetup. Oh, it's in person. Let's see where they're going to be. I always, oh, wait, that's in Luxembourg. So they're tricking us a little bit. I think that's what that means. I think it's the National Video Game Museum. That's what it's called. Hold on. Let's get it in there. You know me. I love my, got to check the venue ahead of time. Oh, it's in Netherlands, but it's called Luxembourg. Well, you know how it is there. You showed that before the map, you know, like how you actually are on the border and the border is not clear anymore. So that's the Netherlands and Belgium. But I think this is going to be awesome. June 4th, Jos Maurice, National Video Game Museum. So that should be fun. It looks like ExaGrid is going to be there. They are sponsoring the event and there will be Jost from Rabobank. I think he's one of our clients and he will give a nice session as well. There's always great networking opportunity with these Vim user groups in person, right? Always good food. And yeah, also playing the museum. I think this is kind of an experience museum. You get the chance to play in there. So what else can you want out of it? I think it's going to be awesome. So good job guys. I just really want to know what is an American buffet. I got to know what that is. What's an American buffet? Does that just mean like cheeseburgers? I think that's what it is. Yes. An American. That's what in Europe, we would call an American buffet. You would have probably... Really? Interesting. Because hamburger came from Germany. Hot dog, sausage was that's European. I don't know what you would call American buffet. Barbeque, like a barbecue. You can Google it just now. So we make sure everyone knows what an American buffet is. We got to settle this. Hold on. What is an American buffet? Now this is giving me self-serving. Okay. I know that. I think what they're getting at is they probably just don't need the word American. I don't know. Yos, whoever goes there, please send pictures. I really want to know. Let us know in the comments, Yos, Maurice, please. Yeah. Good stuff guys. I'm being silly, but most of the time I'm serious. Sometimes I'm silly. All right. Next up in special department. This one was in Arabic, hence the right alignment. I've translated it to English, but Mohammed and Youssef talking about universal CDP. That's great. We don't get a lot of content. Oh, well, first of all, I know that the faces are familiar to you, Rick, but if you look at the banner, it says VUG Saudi Arabia. So this is a new group, right? We are not in VUG Egypt. I called it Egypt, didn't I? This is not Egypt. Saudi Arabia is a new group in town. Behind it is Mohammed and Youssef. Youssef is based, as you know, in Saudi Arabia. He works for BIM and it's great job. Great initiative already 18 people register. But yeah, together we are actually looking to find a local leader, you know, from a partner or from a customer in Saudi Arabia to work with Youssef and Mohammed on this initiative. So if you are watching us and you hear about this and you are interested to be part of this amazing team, please do so. And yes, this is the first virtual event for Saudi Arabia. Super interesting topic, as you said. And yeah, we don't really talk a lot about the CDP and replication, you know, but I think, yeah, we maybe we had it was also maybe Mohammed at some point organizing a few years, one year ago for VUG Egypt, I think talking about CDP, universal CDP. So thank you, guys. Awesome. Looking forward for this one. I'm not going to understand much, but looking forward anyway. Well, I apologize. I missed that. I didn't I didn't notice that we have our first VUG and KSA. So that's great. Thank you, Mohammed. And thank you really for just going out of your territory. That's incredible. And Youssef as well. Thank you for stepping up here on this. This will be great. All right. Going back the other side of the world, I think this has something to do with Esteban, but let's talk about VUG Argentina. I've translated this one to English. This one will be in Spanish in person at the VIM office in Buenos Aires. This looks good. Yeah, for sure. I'm sure Esteban, if he's not busy with a summit or a conference or something, he's going to join the event. This is June 4th as well. So a lot of good things happening around that time. But this time, our Argentina leader, Pablo, put together this event. And it's a team office, as you mentioned. With this occasion, I want to say thank you to Ilana, Tomas and all the team for all their support all the time. They are allowing us to use the office for these VIM user groups. So that's important. And I really like the theme of this one. I think it's good. They always relax, you know, the Argentinians. I like how they do the events. And the theme of this meetup is the culture of resilience, which is very, very interesting. And on top of that, I'm really looking forward to see the guys this year in Buenos Aires. Yeah, yeah, they're, they do it well down there. That's great. Awesome. Yeah, I forgot about Pablo. Yep. He's got to have a hand in this too. Thank you all. And coming up 4th June. All right, this one's fresh off the desk. Badalina, like same day, just two hours ago. Is this a vote or do we have a winner? Rick, with so many good engagements that the community have and so many things happening, I almost forgot, I should have posted it on Monday. And today's Wednesday. And I was like, hey, I did not do blog of the month. And then I had so much trouble selecting six articles, because there is so much good content. It's just getting better and better. So congratulations, all of you. Thank you for contributions. And all the rest that is not mentioned in here, please vote. Help us decide who's going to win that badge this month. I know what I'm going to vote for, but I'm not going to do it while we're recording. Yeah, yes. Thank you. Good luck. Pete, Michael, Felipe, Mateus, Mohamed, and Nico. Good luck. All right. So the last bit of tabs are on me. I want to describe to you, we have an update 13.02 for Veeam Backup and Replication. And we have a couple of patch vulnerabilities that came into our disclosure program. So thank you, Alibos and Parsa, for reporting that. That's a great idea to put the phone on silent. But basically, I want to just highlight that we've taken an update to Veeam Backup and Replication 13.02. And what was fixed in here, we get a list of that. Mostly around components, right? You know, you can see where we started with Antonio's, some of the different individual components, right? Getting updated. So we list out what that is. So 4850.2, and 3103. And then I also want to highlight that Veeam One, as well as the Service Provider Console and the agents, you know, when it all kind of cascades through. So several different updates that I want to just highlight to everyone have been published with 13.02 and more. And believe it or not, Maddie, that's all we got. Oh, my goodness. That's all? Well, I know it's a lot. Well, who's new to me? You're right. That's right. Who's new? Sophia's still on well-deserved vacation. And I want to know, she takes really good vacations. So do you, by the way. Yeah. We need it. I don't know why. We need it from time to time, right? Yeah. Yeah, exactly. So we'll have a catch-up episode on that for sure. But yeah, next week, I'll be in London. I want to see what we're going to do. I'll let you know if I can be on the same time zone. So I might be able to sneak it earlier or later. I'm going to be on vacation. Don't forget about that. Oh, you know what? Okay. I'm just going to say it. I'm going to... Hold on. Hold on. Hold on. You? I'm getting crazy here. I unhooked my audio. Hold on. He's leaving. I don't know where he's going. What's going on? What's going on? So, dearest Maddie, dearest community, I have some news. Ooh, let's see. What are the news? Have you all seen Dalbo? Oh, my God. Congratulations, Chris and Rick. And also, we have to congratulate Nicola because she was writing the preface or what was that? The foreword or the intro. The message. Yeah. So congratulations, all the three of you. Also, Marcus, a.k.a. Dynamic, was one of the reviewers. Aaron Stebbing as well. So thank you both. But the big news, Maddie, no joke, in two different categories on Amazon.com, this is number one bestselling book. Oh, my goodness. Yes. This is reason to celebrate. I don't know how we can celebrate that. Maybe we have to invite Chris to talk about this book in one of our recaps. That is Chris. I think I'm going to have to kind of I'm thinking I'm going to sneak in, Chris, so we can talk about the book. So, Maddie, I'll show you one more thing real quick and community as well. See if I can do this real quick. So I was floored when I saw this. So screen share, screen share, screen share. Are you still having trouble? In the network administration section, as well as the storage section, both sections. Number one. Amazing. Look at that. Yeah. I mean, I don't know how we're going to do it. But if you cannot do it, probably next week is going to be a bit difficult for you. So you might have to just do a solo video or post. But I think we have to definitely schedule one recap with Chris and talk about the book. I mean, it's a big moment. I mean, I've done a number of books and I swore off ever doing another book until Chris said, hey, could you help me? And I'm like, I'll do it. And then it's number one on Amazon. I think that's pretty cool. Indeed. All right. So we'll figure out next week. But if you're on vacation, I want you to vacate and I want you to enjoy it. You know me already. Okay. We'll see. We'll see. All right, everyone. Well, that concludes episode 263. We'll see you somehow, someway.
