Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Database as a service with Percona Everest
0 Plays2 seconds ago
In this episode of the Kubernetes Bytes podcast, Ryan and Bhavin sit down with Edith (Edi) Puclla, Technology Evangelist at Percona to talk about Percona Everest. The conversation focuses on Percona's investment in the Open-source ecosystem, and how they keep innovating with Percona Monitoring and Management and Percona Everest. The discussion also dives into how the community is expanding and supporting more use cases with Data on Kubernetes and AI on Kubernetes.
Check out our website at https://kubernetesbytes.com/
Cloud Native News:
- http://businesswire.com/news/home/20250227229731/en/Veeam-Releases-Kasten-for-Kubernetes-v7.5-Elevating-its-1-Data-Resilience-Market-Position-with-Enhanced-Security-and-Modern-Virtualization/
- https://www.developer-tech.com/news/kubescape-achieves-major-milestone-in-open-source-kubernetes-security/
- https://thenewstack.io/postgresql-operator-cloudnativepg-hits-the-cncf-sandbox/
- https://backendnews.net/tenable-report-finds-security-gaps-in-kubernetes-environments/
- https://www.infoworld.com/article/3832760/red-hat-openshift-improves-virtualization-support.html
- https://edera.dev/stories/series-a - https://www.datastax.com/blog/ibm-plans-to-acquire-datastax
Show links:
- https://github.com/percona/pmm
- https://github.com/valkey-io/valkey
- https://github.com/percona/everest
- https://github.com/percona/percona-server-mysql-operator
- https://github.com/percona/percona-server-mongodb-operator
- https://github.com/percona/percona-postgresql-operator
- https://forums.percona.com/
- https://github.com/percona
- https://www.cncf.io/blog/2024/06/28/celebrating-10-years-of-kubernetes-the-evolution-of-database-operators/
- https://dok.community/
- https://www.linkedin.com/in/edithpuclla/
Timestamps:
- 00:07:40 Cloud Native News
- 00:17:40 Interview with Edith
- 00:58:05 Key takeaways
Recommended
Transcript
Welcome and Podcast Introduction
00:00:03
Speaker
you are listening to Kubernetes Bytes, a podcast bringing you the latest from the world of cloud native data management. My name is Ryan Wallner, and I'm joined by Babin Shah coming to you from Boston, Massachusetts.
00:00:14
Speaker
We'll be sharing our thoughts on recent cloud native news and talking to industry experts about their experiences and challenges managing the wealth of data in today's cloud native ecosystem.
Hosts' Return and Personal Adventures
00:00:30
Speaker
Good morning, good afternoon, and good evening, wherever you are. We're coming to you from Boston, Massachusetts. Today is February 28th, 2025. Hope everyone is doing well and staying safe. Yes, you heard that right. It's 2025. It's been a while since you've heard our voices, but Bob and I deserve a break ah as well. Okay.
00:00:49
Speaker
yeah yeah for sure What have you done doing it all this time, Bhavan? Did you go anywhere exciting, do anything fun, are you just hunkering down and hoping for warm weather these days? I did a... Okay, when you asked me the question, i was I had to think about it. So I did a short trip to Italy over the President's Day long weekend. nice.
00:01:09
Speaker
I spent four days. and What did you do? do ah Went to Florence and then Rome. I think Rome was a lot of fun because, again... Again, I'm not a Christian, so I don't know about all the history of the church and Pope and things like that.
00:01:23
Speaker
But visiting Vatican, doing like a guided tour that helped. And then ah Roman history as well, like new bits and pieces of it, watch the gladiator movie. But doing the old one or the new one.
00:01:35
Speaker
ah the I've seen both, but the new one sucks. I don't think. Really? i was pretty excited about seeing it, but now you just put a real big hankering. I mean, the original was so good. And the first one was really good.
00:01:50
Speaker
Yes. Like I watched the first one again before watching the second one. And then I was like, what are they doing? Like how did they spoil such a ah good good movie? but yeah I'm bummed.
00:02:01
Speaker
yeah You should still watch it. like they have an exciting well but I'll probably just wait until I can like you know rent it on Apple TV or something. So the the Coliseum tour and the Palantine Hills and Roman Forum, all of those things, that tour was awesome. like Man, that the tour guide that we had, he spoke for four hours straight. like I started listening to him. I don't know how he kept doing And I'm sure he does this at least once or if not twice a day and and says the same thing. like I don't know if I can do that job, but no. Did you go in the leaning tower? did you go over to ah Pisa?
00:02:36
Speaker
no Nope. No, I didn't make the hump. Yeah, I didn't have enough days. oh That's a bummer. I've heard it's continually like... leaning but maybe that's maybe that's just hearsay or a rumor I just built that way right like it's not that I know but I've i've heard that it's like i mean it could be totally conspiracy yeah that I've bought into but I've heard that like very minute amount that it's still leaning Oh, it's still going?
00:03:03
Speaker
I guess. I don't know. i got People going on top and jumping so that they can use gravity to help the cause. um It's probably not true, but I like to think that it is because it makes it more interesting.
00:03:16
Speaker
No, so that was a fun trip. How about you, Ryan? Did you eat anything anything good before we move on? Yeah, for sure. Like, I love the pizza. Like, I know it's like, yeah, not a surprise, but yeah, it's it was really good. Like, you get these really thin crusts so squares, I guess you can buy buy them ah yeah on on weight rather than slice or a pie.
00:03:38
Speaker
ah And even for vegetarian pizzas, man, the sauce, the the bread, everything was really good. Like, yeah, would recommend that. um i'm I've never been Italy, so... You should, man. Like, it's easy. Like, man, there is a... Apparently, there's a direct flight. Like, I had to do a layover, but ah I don't know. Super easy to get there. ah Language barrier wasn't really as much. Like, okay, i so I was doing Duolingo.
00:04:02
Speaker
Had a streak of, like, 100-200 days trying to learn the language a bit so I can have basic conversations. I figured out how to say, hey, how are you? Can I get a table for two? But I didn't realize this...
00:04:13
Speaker
ah what do I like? I don't know what they are saying in response. You can't understand them. even They think you're speaking and then you're just like, what's all I got? yeah And then i was like, ah yeah, sorry. Like, I know you said a few sentences, but I didn't get any of it.
00:04:29
Speaker
i was I was just trying to to to to help. I mean, i think like you're right I think it is appreciated that you're trying, even if but even if it's terrible, right? um Yeah.
00:04:40
Speaker
So next time I need to learn what the response would sound like. Or just have your Google Translate open when they respond or something. That's true. Yeah. But I hate Google Translates. um So for text, it's awesome. For just basic words, it's good. But if you want to take a picture of a menu and then translate it In the past, I've had a good experience, but not in Italy. Like it just, I don't know. It never gave right. That's fun. You just kind of like, this one looks like it'd be good. Yeah.
00:05:06
Speaker
Yeah, that's true. Chef's choice, please. Yeah, whatever. Yeah, just just for me, no meat, please. That's it. Yeah, yeah, exactly. You'd have to say something like that. I'm i'm a very, like, i like experimenting. So I would probably just say, like,
00:05:23
Speaker
you know, menu roulette just yeah there. Well, i did not go to ah Italy or anywhere as cool. I went to Vermont a couple of times, which is great. Did some fat lighting. It is not cool. It is cold. It is cold.
00:05:36
Speaker
I was, i took a picture next to up in, up in Vermont. I think it was the beginning of February next to this icicle that was like as tall as me ah because it was that week, you know, a couple weeks after it was like two degrees or whatever. Like,
00:05:52
Speaker
It was fun, though. I i did a couple different um fat biking trips, which is always a blast if you like that kind of thing. or Or something, you know, it sounds terrible to some people, mountain biking in the winter. But one was sort of an event, and the other one was just like a ah two-day trip.
00:06:10
Speaker
But yeah, no no Italy, that's for sure. Do you have any tips? that you won't talk about i need no more trips planned really um no uh i'm going back up to new hampshire this summer as always but nowhere super fun i've been trying to figure out uh how to get down to uh florida in the winter that might happen in april but that's not even it's like spring break that's true it's It's starting to get warm enough here where I'd be happy.
00:06:42
Speaker
yeah i think if you're to do that, it's going to be in the middle of ah winter. February, I guess, yeah. Last year, I did the same thing. like I went to Florida for like three, four days because I need a break from all this cold. I'm done. went By the end of February, I'm done. i'm like i like winter. I generally like winter. I went snowboarding a bunch, you know fat biking, but I'm like, I'm ready now.
00:07:01
Speaker
need some... yeah Yeah, please. Freeze, dirt and grass and warmth yeah and sun. Please. Bring the 40 degrees, 50 degrees weather. Enough with the below freezing. um um Yeah, I'm done.
00:07:12
Speaker
Done. We did get, I don't know, a good amount of snow this year, right? Like Boston got like 28 inches this winter already? Yeah, definitely. Well, I mean, compared to last year, it was, yeah mean, any any snow would be better. It basically just rained all last winter, which was kind of nice, actually. Yeah.
00:07:29
Speaker
All right, enough about weather. um We have a fun show ah today. um We'll introduce our guests in a little bit, but we are going to dive into CloudNave news. ah We're not going to cover all the news we missed.
00:07:41
Speaker
That's just not what we're going to do because no one wants to hear that much news before an episode. But we are going to kick off... um with a few new stories from the both of us. i got I got a few more here, so I'll start, I guess.
00:07:55
Speaker
um The first one I have is um around Red Hat. OpenShift 4.18 is the latest version of OpenShift. And the main thing there that I wanted to kind of ah note, which is um definitely been bigger news from the start of 2025, is their OpenShare virtualization stuff is is GA and some of their networking stuff that supports A lot of what they're doing there is GA as well.
OpenShift 4.18 Update: What's New?
00:08:28
Speaker
And this kind of just, I think it solidifies to me, right?
00:08:32
Speaker
um ah You know, when when they're bringing things like VM live storage migration to GA as well, um supporting more and more virtual machine use cases, everything with Broadcom that we talked about last year, it's coming to fruition now. Everybody's trying, you know, I shouldn't say everybody, but a lot of people are trying to find an alternative OpenShift virtualization is a big one on top of, ah you know, what SUSE has got going on and and a number of other ones.
00:09:00
Speaker
um So I think exciting stuff here. um You know, if you if you haven't been following what OpenShift is up to in OpenShift virtualization, definitely go take a look. This article will kind of, you know, give you a little tidbits of of all the stuff there they're kind of thinking about.
00:09:18
Speaker
um You know, there's there's so much in the VMware ecosystem that, you know, you know, people are going to want, and it's, it's so early in the OSV and QVirt world where there's going to be gaps for a while, but you know, they're making strides.
00:09:34
Speaker
Yeah, and I completely agree. like We have had, what, 20 plus years to build the features that they had. oh kind like Even though Red Hat moves fast, but two years is, like we have been hearing more and more about OpenShift virtualization in the past two years. I know Qubat has been a project for longer, but it takes more than two years to get and all the features that customers need. So I think they are making great progress, right? like They are building the right features that are more relevant and then add other things that might be a nice to have, I guess.
00:10:01
Speaker
So, yeah, I mean, even some basic stuff like the article I have talks about user defined networking, right? edns And, you know, just getting things like when VMs move to making sure IPs don't change. And, know, these are, i would call basic needs of, of these types of use case use cases, but um I mean, the solution works really well. I've, I've been having my hands on it recently and it's, it's,
00:10:26
Speaker
It's very capable. I'll put it that way. but It does what it's advertised. If you're a V expert, you you're going to know a lot more and and you're going to probably find a lot more gaps. Yeah, true.
00:10:38
Speaker
All right, over you. Oh, okay. So I had ah one fun new funding round. I know the the market, man, it's crazy, but I'm glad that Kubernetes or cloud-native focused companies are still getting funded.
00:10:51
Speaker
Adira is a startup that I'm talking about. We we did discuss them when they raised their seed funding of $5 million dollars back, I think, in October timeframe. They are now following that up with like a $15 million dollars Series a Seems like a quick turnaround, but again, I think ah they are focusing on an interesting use case where they want to deliver like a security solution that makes cloud-native installations more secure by having better isolation.
00:11:17
Speaker
know when we covered their seed round, we were talking about how they want to build a new Zen-like hypervisor using Rust. So I think that they're making good progress there. ah They also have another product called Adira Protect AI, which is, again, something that that's new, right? Maybe have a few customers ah using it or testing it out, but They market it as like industry's first secure by design solution that automates GPU configuration. So I guess bringing in some of those isolation capabilities
Funding and Market Trends in Cloud Security
00:11:43
Speaker
in GPU. So in their ah funding blog, they did mention that um the vulnerabilities that were recently discovered in the NVIDIA container toolkit
00:11:51
Speaker
If you have something like some a solution so which is similar to theirs, your GPUs are already isolated. So even if you are vulnerable, ah you might not be exploitable. So I think they're going after a good space, obviously tying it into an AI use case and talking about GPUs and NVIDIA that obviously gets you money, but a good startup in our ecosystem for sure.
00:12:11
Speaker
Yep. Fun stuff. um The other story I had was back around security. So tenable security findings specifically. This is another ah cloud risk report. This was actually done in 2024, but the article is from February 2025. It goes into a lot of the the things I think we talked about last year in our security episodes, which was things like misconfigurations, ah publicly accessible API servers and Kubernetes.
00:12:43
Speaker
um you know, ah I am credentials having excessive permissions, ah cloud storage being publicly accessible applications having, yeah, having and unremedied cve So like, it's a, it's a real challenge I think is really just what this is calling out. And, and I think,
00:13:07
Speaker
especially in 2025 and beyond, we need to be better. um So just, you know, this article is more of a learning, um you know, ah resource around some of those common vulnerabilities. Maybe you can go back and take a look at everything you're doing. I know I've i've been working on an application that I have, you know, depend about telling me when things are,
00:13:32
Speaker
are out of date or have security issues. And it happens all the time. You really have to keep up on it. So um definitely go take a look. um The other three I'm not going to go into in detail, but um they're more about new releases. So Veeam 7.5 is now out, ah has some support for Veeam.
00:13:51
Speaker
you know, virtual machines and Kubernetes, which a lot of the Kubernetes native, you know, solutions are now making sure they can ah handle virtual machine workloads that are basically, um,
00:14:04
Speaker
containerized in some shape or form um and be able to move those around. CubeScape from our friends at Armo is now an incubating phase and CloudNativePG from EDB ah is now a sandbox in the CNCF. So some fun stuff going on there.
00:14:24
Speaker
That's awesome. I think for for me, I know we covered the funding round. We also have an acquisition that is kind of in our ecosystem. Yeah, that's the one that a lot of people are talking about. Oh, yeah. And and again, ah I wasn't expecting this. I and know before we hit recording ah hit record, I was telling you that I didn't expect data stacks to be looking for an exit, honestly.
00:14:46
Speaker
So I was surprised. Yeah, wasn't on your bingo card, according to you. Oh, yeah. Somebody is keeping up on LinkedIn. Yeah, so data stacks got acquired by
IBM and DataStax: Strategic Acquisition
00:14:58
Speaker
IBM. So data stacks, I know we have had Patrick on the on the podcast earlier.
00:15:02
Speaker
ah to talk about Cassandra and their are Astra, which is a managed Cassandra service. If you were keeping up with them, so in addition to them being a pioneer or a leader in the Cassandra ecosystem, they also started to ah ah going after vector databases and building Langflow to help ah developers build AI-based applications and use Cassandra as a database.
00:15:22
Speaker
So they have been doing some great work. So IBM is is definitely surprising all of us, right? Like with Red Hat acquisition, that's doing really well. ah Just yesterday, they closed their HashiCorp acquisition as well. And now ah they announced acquisition of DataStacks. So obviously they're building a good portfolio of companies under them.
00:15:42
Speaker
yeah It would be interesting to see where all of these fall. Do they still kind of operate in the alphabet model where each client team or each company um still acts as its own entity.
00:15:53
Speaker
ah But no, I think, yeah, IBM might not have been the place to go or place to work for. But now with these companies, they're definitely turning that around for sure. At least for me. Yeah. And their GM, Ritika, i think, ah you know, the blog post by her, right? There's there's some other things in here that are tailored towards AI too, right? um So DataStax integration with Watson, I think is is planned, but also the fact that DataStax is the parent company of Limeflow, the open source project to orchestrate AI applications. So I think there's sort of a you know technology acquisition there as well for for a lot of things. IBM has been sort of...
00:16:34
Speaker
um you know focusing on in the AI realm as well. That's such a great point, dude. like IBM with Red Hat already has a set of LLM models, right? They call them the Granite models ah that ah there that they have open sourced. And now with Limeflow and DataStacks and Cassandra-based database, they have, of oh man, could this strategy makes sense. I'm excited now.
00:16:54
Speaker
but ah you You think so? Yeah.
00:16:58
Speaker
ah But yeah, fun stuff ahead for the Data Stocks folks, and congrats to all those there enjoying the acquisition. Cool.
Guest Feature: Edith Pugja from Percona
00:17:06
Speaker
um Well, we have a really fun guest, um a fellow DOK ambassador, Edith, ah from... um Percona. She's a technology evangelist there. And um we've had the pleasure of interacting with her and the DOK community and and and generally in the Kubernetes community. She's a lot of fun.
00:17:28
Speaker
So she's going to come on and talk to us about um ah what's going on at Percona and everything they're doing in the open source land. So without further ado, let's get you the thon.
00:17:38
Speaker
All right. Welcome to Kubernetes Spites. Edith, so glad to have you on the show. I know we tried to get this done last year, but we're doing it now. So welcome and please introduce yourself for our listeners.
00:17:50
Speaker
Yeah, hello, Bahim and Ryan. Hi everyone, thank you for having me also. I'm Edith Pugja, I work for Percona as you know as a technology evangelist. I also have a role as a QA in the Everest team, Rasion Tli, so I'm really enjoying.
00:18:07
Speaker
and So Percona, for people who don't know, Percona is an American company that provides services. and open-source software for databases. so Our main focus is databases. Our goal at Percona is to help our community and our users to run databases faster, better, reliable, and now on Kubernetes.
00:18:29
Speaker
Yeah, yeah, of course. And yeah there's a trio of DOK ambassadors on here. i feel like this is fun. So data on Kubernetes is near and dear to our hearts. So if you're ah if you're listening and interested in there, ah you'll you're in for a good episode. um and And glad to have you on here. So, you know, ah you mentioned Everest and we'll get into that in a little bit.
00:18:51
Speaker
um But I do want to start with sort of the open source side of Percona, right? um I know we had talked about this a little bit, but maybe you could answer what does open source mean in the context of Percona and what you do dayto- day to day?
00:19:07
Speaker
Yeah, and for Percona, ah open source means sharing software that anyone can use, can change, and can improve for free. We know that. Open source is about ah also working together as a community openly, giving people the opportunity and the freedom to customize their databases, in our case, because that's our focus, without being tied to one specific company.
00:19:33
Speaker
Also, Percona believes that open source helps everyone um by letting people share their ideas and improvements, which is other in open spaces like GitHub or like, yeah. yeah um Yes, this is what what we do. And we are constantly also asking for feedback yeah for people in our community.
00:19:53
Speaker
We have our forum where people can discuss of ah about our open source projects. We have also GitHub. which is open to discussion and feedback always.
00:20:05
Speaker
Gotcha. And Edie, I think we had Peter, ah the one of the co-founders of Percona on the podcast, like maybe three years back at this point. And yeah that's since then, we have been following Percona and all the great work that you guys do. right ah and but So I wanted to focus like more on the recent times in the cloud native ecosystem or the Kubernetes ecosystem.
00:20:25
Speaker
What are some of the open source projects that people might be familiar with or are working on? I know you have a Valky based solution as well, or you guys are a contributor to the Valky project. Can you talk a bit more about that?
Valca Project: A New Redis Fork
00:20:38
Speaker
Yes, ah most of our projects are ah i know now, are they are trying to get focus into clonative. ah We have from our ah standard projects that I can say, for example, Percona Toolkit, the database administrator to use, or PMM, but they, internally, they can use they can be used also in clonative projects.
00:20:58
Speaker
For example, we have our projects ah for Kubernetes operators. We have Kubernetes operators for different databases. PostgreSQL, MySQL, and MongoDB.
00:21:09
Speaker
So they work on clusters of Kubernetes. um ah After that, we have also Percona Everest, which we're going to talk a little bit about that after.
00:21:19
Speaker
But for Valca, this is one thing that ah ah is very is very interesting. This has a very interesting history also. It started in March in 2024 after we know Redis decided to change the the license ah of the code. it become and a in property So as a result, ah the community also and the developers, they decided to fork that the the repository and they renamed it as a. And they they are keeping the license as a BSD.
00:21:51
Speaker
We know Bergeley software distribution license. Then after that, the Linux Foundation take the control of this project, allowing the community to send changes, being half-maintainers, contributors,
00:22:04
Speaker
And to they this community is still growing. um We as a Percona, we are big support of this project. We have maintainers also inside Percona. And the co-founder of Percona, Peter is one of our co-founders. We have Badin Percona.
00:22:20
Speaker
Kachenko, who is also a co-founder of Percona, he's leading this project of Valcay for Percona. So we are sending improvements, we are improving the functionality and that help companies also migrate to Redis. um As a company, we are also helping, a giving support for people who want to migrate for Redis to Valcay.
00:22:41
Speaker
So we are 24 for 7 supporting that kind of migrations. Yes, it's what we do. So in in March 2024, was this like a ah fork of Redis before after they changed their license? ah okay Okay, so it is starting from the Redis code base.
00:22:57
Speaker
yeah It's just now an open source. Yeah, okay there was a big thing. I remember the last year people were talking, wow, that's happening. Oh, how is this possible? yeah um But fortunately, the community just was aware about that and they just forked the project and then, okay, what we are going to do?
00:23:12
Speaker
and all the process became right. Then the Linux foundation started and all the things. And now we have an open source project, and we are still making better. A lot of companies are getting into that and getting more popular.
00:23:25
Speaker
It's going to be, I guess, a wait, it was, yes, there was, a they are having also boots in several lab events. it oh wow so There was a SOCOM conference, for the State of OpenCore here in London, and they had also booths there. So they were they were talking about what is this about because not much people know that there is still in an open source project for for Redis.
00:23:48
Speaker
ah No, I think now the community is so proactive and I guess quickly reactive. Whenever there are licensing changes, like we saw that with with Terraform and OpenTofu as well and then Redis and Valky.
00:24:02
Speaker
I also learned sometime back that OpenSearch is nothing but Elasticsearch. So like all of these things are just new names and open source versions of existing things that people might be comfortable using. So no, ah thank you for sharing the the history on Valky.
00:24:16
Speaker
Yeah, and we'll um you know we'll include some links that Edie gives us on Valkyrie if you are interested in that project specifically or have more questions for her as well. So um I'm going to switch gears a little bit here, Edie, and talk about little more about running databases on Kubernetes. and And one of the first things I want to ask you since you were ah role at Percona, you do various things, is you know what do you feel are, you know maybe today, what are the challenges that organizations have, or maybe the most common ones, running databases on Kubernetes today?
Databases on Kubernetes: Challenges and Solutions
00:24:48
Speaker
Yeah. um I was ah thinking about this this the last time because I had a presentation yesterday about Kubernetes operators. And i was asking and developers in my team, hey, what's, what do you more struggle? Because they are the one who daily a support this kind of activities, right? Of people running databases on Kubernetes.
00:25:11
Speaker
And there was something interesting that i they found that is a that ah the big challenges that they found is that organizations that run databases on Kubernetes is that complexity and it requires a specific skills for running databases.
00:25:28
Speaker
I mean, the just the knowledge that we have right now as ah maybe for Kubernetes, basic knowledge and even intermediate is not enough. We need more because we are talking more about um advanced but processes for databases.
00:25:43
Speaker
no And ah so many teams don't have enough people with the right experience, which make this adoption complicated. Some of this is because of the lack of technical knowledge, knowledges as i as I said.
00:25:56
Speaker
But sometimes this is also for the hesitation to ah start to touch something new and also complex. Right. Yeah. I i myself, are I'm included in that. yeah Yeah. the good So this sounds like it's it's more like Kubernetes has done a really good job or the community in general has done a really good job, including operators, getting to a point where you know deploying them and getting them running on Kubernetes is is is maybe a small percentage of really what you're going to have to do over the lifespan.
00:26:29
Speaker
um and so This sounds like more like day two operations and beyond. like How do you manage these things at scale? and and Those are things that DBAs have been doing for so long, but maybe Kubernetes in our community, which focus on on sort of operations,
00:26:43
Speaker
Does it have that ah skill set as much is kind of what I'm hearing. Okay. Yeah. the The good thing, the the good news ah is ah that this started to change as you just mentioned it, right? The communities exponentially growing exponentially.
00:26:58
Speaker
A lot. I saw yeah ah saw communities, a events, meetups, a Kubernetes Community Days talking about bringing this community ecosystem, cloud-native community ecosystem, to places that we cannot even imagine.
00:27:12
Speaker
in In the the city where ah where I work, we did lots of a lot of resources. there are three two There are two meetups running about in Peru. There are running events for communities. so I was amazed about how everything is growing. Even you throw Kubernetes bites, you are bringing this kind of things to people. So the community is just growing a lot. And it's the way how how we are mitigating this ah this risk of of knowledge, right?
00:27:41
Speaker
And as you tell also that operations that ah companies does in the day after the day one of operations, it came to day two of operations, right? Day one is everything is up and running, our database is running, but after that, what?
00:27:56
Speaker
There is a lot of stuff that we need to do e scaling, backups, storage, failovers, replications. Oh my God, there is a lot of how we can handle this. So there is another challenge that also Kubernetes operators is is helping. yeah So Edie, I have a follow-up question there, right? So in in the past, let's say in the non-Gubernetes world,
00:28:18
Speaker
I remember like early in my career, there were dedicated database admin teams. So there was a SQL Server expert and a team that worked for him, an Oracle expert. And you relied on these guys to know all the level all the level of details, all the intricacies of how how things work, how they should be backed up, where the log files should go, how frequently they should be flushed, things like that.
00:28:40
Speaker
So with the with the community outreach that you just mentioned, right like the local events where people are using those, participating and upskilling, do you see people who started with Kubernetes now being more ah trained on database technologies?
00:28:55
Speaker
Or do you see that the community itself is expanding and all those SQL Server experts or Oracle experts are now learning more about Kubernetes and how they should change the way they operate their databases because of this new technology stack? Which way is the transition going?
00:29:11
Speaker
um I see the necessity a i see the necessity that people that work with databases are seeing now, and um they see that there is one technology that they can get to they they can help to automate Sure. There are things that I do manually.
00:29:31
Speaker
The technology is there, but the step to start to take it is what is changing these days. they see or they They listen about Kubernetes in all these meetups and all in the community, but they say, okay, the technology is there, so maybe I can use it for a Kubernetes. But no, Kubernetes is used for a stateless.
00:29:47
Speaker
No, I'm taking my work. But no, it's also for a stateful set application. So I can i can't try to do it with my database. But here is the question again. They say, OK, should I jump just because everyone is jumping?
00:30:02
Speaker
Yeah, sure. Sometimes we have to evaluate. It's a fair question. It depends. It depends. Okay, is you have the knowledge to jump into Kubernetes. So the amount of data or the transactions yeah you you are doing is good to jump into Kubernetes. You have to analyze yourself if you have all these components. So then, okay, I need to scale.
00:30:21
Speaker
i my my not my My application itself, my business itself requires that I need to scale. So maybe it's a good idea to jump into Kubernetes. Yeah. yeah And I know you know there's there's always going to be a line to where Kubernetes sort of ends and operate you know the database ah admin sort of begins again. It doesn't mean the database doesn't need to be administered just because it's running on Kubernetes.
00:30:45
Speaker
yeah Some of the operations get you know somewhat replaced, like how you configure clustering or replication, those kind of things. but There's still someone that has to sort of you know keep an eye on it, um which you know the metrics and stuff like that can come out of Kubernetes, but still someone has to know what they're looking at or manage schema migrations. you know i don't think Kubernetes is going to do that for you.
00:31:04
Speaker
Lots of other things. yeah as well i don't so get yeah Go ahead. Sorry. One more question, right? um the The traditional databases, I know when we talk about Kubernetes or modern cloud native distributed databases, we always talk about ah Cassandra or Postgres or or Mongo or things like that. like is there a train Is there a shift that you see in the ecosystem that, oh, yeah, I might be an Oracle expert, but now I know that my application developers or my application teams need me to be a Cassandra expert. So are they also changing the way they
00:31:37
Speaker
changing the the databases that they were ah they used to own and deploy as well.
00:31:43
Speaker
I'm not sure about that. ah but What I see is that, for example, PostgreSQL is the most favorite ah database a and it is still being the most favorite database.
00:31:55
Speaker
It means I don't see too much that that chain between that. grace um yep ah Yes. I'm going back also into into your comment, Babim and Ryan. and I don't see it as a Kubernetes operator or something that could replace what ah what ah database people yeah does, but it helps a lot. So database ah people or the team could focus more on what you said, Ryan, about monitoring what is after, okay, I should keep the database alive, monitoring, observability, all that kind of things.
00:32:25
Speaker
Yeah. Yeah, and i think in general, it's it's definitely more common, or least for for what I've seen as well, is when you have ah an organization working with multiple databases, it could be multiple of the same kind, right? It could be you're running several instances of of Postgres um or even different types of of databases, as you mentioned.
00:32:46
Speaker
you're going to find more of those use cases in Kubernetes than just like a one-off database. But you you mentioned monitoring, and and this is a good lead-in for the next question is, you know how how are you seeing organizations ah monitor the databases um when it comes to
Monitoring Databases in Kubernetes
00:33:02
Speaker
Kubernetes? And then maybe talk about how you see it in the Percona world as well.
00:33:08
Speaker
ah Yeah. m About monitoring, I saw that more people are taking this to the next next step. There is not just about um knowing what your database here is doing, but there is more critical thing now that there are more tools available to do it, to do observability. And even now people are are doing and adding features to it with AI.
00:33:30
Speaker
So this is absolutely the next a step for that. But I see that ah people have a very um um responsibility for that, for observability. or there are different tools that they they can use for getting even alert alerts in real time to know what the data ah ah what the data what the databases are doing, who who is accessing to the database, also to have real-time alerting and warning, OK? And they can fix before these problems happen, right? They have real-time visualizations through the graphical ah ah dashboards. I don't know you've seen this dashboard.
00:34:09
Speaker
They are amazing because they give you a lot of information. I don't even understand all of them. There are metrics. ah queries, how your queries are going, which databases are getting a lot of queries, for example, for potential attacks.
00:34:21
Speaker
All that things and vulnerabilities ah you can see in in your m ah monitoring tools. um Yes. So you you brought up AI, right? Like how are how is AI affecting these monitoring capabilities? Like are people moving to natural language-based queries on their environments and asking how the databases are doing or they're still going to those dashboards and looking at at at those detailed metrics?
00:34:48
Speaker
I was really curious about that also in some of the events. And I was wondering, but I was reading that people are in this stage, they are in in testing period, I can say. sure They are trying because they don't want to jump into AI and start to take and it say, let's do it.
00:35:03
Speaker
Imagine they have some issue and you, oh, you forgot or false positive is what they call it, right? yeah positive And you just ignore that alerting. That could be a big vulnerability.
00:35:13
Speaker
They are in this state. They are just investigating. They are checking. And I think they are seeing what what is worth to have it. But I don't see something that, I didn't see something that is, okay, this is AI is helping in this way and we are going to start to use it from now for this, for this.
00:35:32
Speaker
But for example, if you have a lot of luck for troubleshooting, You can use AI to get which one in in which part I can get these vulnerabilities and how I can fix it. that And you can align that one with your guidance, the standard documentation that you have to do a match and you can have. you can In some way, you can reduce the time consuming that you were having before. Yeah.
00:35:58
Speaker
yeah Those, yeah like asking the ability to ask a AI questions about my, I don't know, 10,000 line log file is definitely helpful. So I think one more question on this, right?
00:36:09
Speaker
um Do you see people being more comfortable with vendor specific AI solutions. So like if I'm using a specific monitoring tool or observability tool, use an AI chatbot there to ask questions about my logs or take all of that log, put it in like an open AI or a Gemini ah prompt and then ask it questions. Like where do you see this happening? Like vendor specific solutions or something that's more centralized that can answer all kinds of questions, not just database related questions.
00:36:39
Speaker
That's a good question. I don't have a visibility for that, to be honest. Yeah, I'm not in that step yet. But I can see as ah I can put in that place, and I can put in the place of from people and I would like to have some centralized things where I can. i can yeah Yeah, I can. More than one thing. Yeah. Yeah. yeah But for testing purposes, yes, then.
00:37:00
Speaker
I think in one year we are going to start to say, hey, this works, this doesn't work. So we are going to start to replicate. So are you saying you need to come back on this podcast in a year? Yes, we will. You know, it's going to be different. and We are not, maybe, it's going to be totally different.
00:37:14
Speaker
yeah Yeah, and I feel like a big differentiator today is you you sort of, as an individual using AI, you still need the background knowledge in order to understand that you're you're possibly getting a false negative or or know enough to look into what it's telling you.
00:37:28
Speaker
yeah versus when you can go to an AI solution and not have that skill set yet and and and get actionable information out of it because you you don't know how to double click yep and kind of dig into it. That'd be the, i think, maybe the the next ah phase once we can fully trust it. But um I do want to give ah ah ask one more question the monitoring side of things.
00:37:51
Speaker
Now, um Kubernetes gives you monitoring solutions out of the box, but um we did mention Parcona monitoring and management, which is another open source observability project from you.
00:38:03
Speaker
yeah How does that differ from what you get out of the box from Kubernetes? Yeah, Percona Monitoring and Management, PMM also, is very popular. I was surprised because a lot of people, when I go into talks, they know about it, they are using, and they have positive feedback about it.
00:38:20
Speaker
It's open source completely, and it lets you monitor databases like MySQL, PostgreSQL, and MongoDB. um It offers detailed performance metrics, query analytics, and tracking is low when your your queries are slow.
00:38:36
Speaker
and notify with a a database ah potential issues, for example. I saw that people are more using open source also because they have the flexibility to change and it is not difficult to install. I mean, the documentation, there is a lot of documentation there and that people can just can just go and jump and do the solution by themselves.
00:39:00
Speaker
But sometimes it's a little bit difficult ah to to go by yourself. So it's where Percona we but gonna re enter into action and we have um support for people who are not able to to do into the next step. Maybe if we start Percona monitoring and management or this kind of tools.
00:39:19
Speaker
So we have we have we have we help, we are helping with that kind of support. But what answering your question, Ryan, I see that um ah people like flexibility.
00:39:31
Speaker
They don't like to be tied to a specific company. They want the flexibility to move from one cloud to another cloud or to hybrid or maybe to your own premise. So open source ah give give is giving you that kind of flexibility.
00:39:46
Speaker
Gotcha. Okay, so I think but one question about Percona monitoring and management, and hopefully that is a good segue to my next question is, is Percona monitoring and management ah a multi-database solution? Like, can it do do multiple databases? And having, and that being a successful open source tool out there, is that what led you guys to build Percona Everest? Like, okay, let's do more than monitoring.
00:40:10
Speaker
Yeah. Yeah. week i We can have three databases, so the most popular and even more. I'm not sure about the more databases that we support. I think MariaDB also is included and other databases.
00:40:22
Speaker
I have to check but yes that. But Percona Monitoring was a big success for us. And after that, um there are many open source solutions. And that is what I like about Percona.
00:40:33
Speaker
The mindset is open source, always open source to the community, right? And this is how also... they started to create Percona Everest. know awesome So if if we are in an elevator, not recording a podcast, can you give me like ah an elevator pitch of what ah Percona Everest is, please?
00:40:51
Speaker
ah Okay, look, Percona Everest, you will see, and you will be amazed using it Percona Everest is a tool that you can install to monitor in your databases. What databases?
00:41:03
Speaker
MongoDB, PostgreSQL, MySQL, and you can have in real time everything that you need to check if your databases are running correctly. And you can solve the problem seeing just that kind of trace and metrics that PMM is going to show you.
00:41:20
Speaker
Gotcha. So instead of all different database admins needing to learn their own deployment tool or deployment operator, Percon Everest gives me that one single unified operator or one single tool to deploy all solutions.
00:41:33
Speaker
Yeah, that you can connect your databases with PMM and you can have the visual visualization of all your databases in one single dashboard. So you can also customize your dashboard. You can customize and say, okay, this is not usable for me. This is useful for me. So I will take it. and Yeah.
00:41:52
Speaker
Makes sense. Makes sense. So, um you know... For Kona Everest, it is it an operator or does it also deploy operators that know how to deploy those various databases?
Percona Everest: Database Management Tool
00:42:04
Speaker
Yeah, you you say For Kona Everest, since we started to work to get into cloud native solutions, we started to build our operators.
00:42:16
Speaker
We build operators for our three databases, for the three one, okay? Operators for MySQL operators. We call it Percon operator for MySQL, Percon operator for Postgres, Percon operator for MongoDB.
00:42:28
Speaker
Okay, there are focuses in our own solution. Percon Everest has his own operator. This operator acts like a orchestrator of the other operators. If you see architecture of Percon Everest, you can connect ah as a user, you can connect through the CLI, you can connect with a graphical user interface, or even with a with CI continuous integration, continuous delivery applications, you can connect with Percona Everest and Percona Everest inside is managing these operators for each database and the operator is managing the database.
00:43:01
Speaker
This a inside a Namespace and inside a Kubernetes cluster. So in one Namespace, you can have three databases, three operators. You can have in another Namespace, you decided you can have also just one database or it depends on what do you want to achieve.
00:43:19
Speaker
Got it. And now, um in terms of managing those various operators, does Everest also handle things like updating those? Or or how does, you know, if i'm if I'm the person who's sort of in charge of, you know, deploying Everest and and and making those available for my team, um how do I go about sort of making sure, you know, Everest and the various operators are up to date? Or does Everest help me do that?
00:43:43
Speaker
Yeah. and If you want to see what kind of operator you are using the version, you go deploy Everest, you enter into the UI, and you will see what version of each operator are you using. It's going to show you, okay.
00:43:58
Speaker
It's going to also show you, okay, your your operator is not upgraded. So you need to upgrade. It's going to give you the option there to upgrade, but it's going to give you the option to upgrade a specific operator, not all at the time.
00:44:10
Speaker
You choose what operator you want to upgrade. Okay, makes sense. Okay, go ahead. dry but but All right, so you know ah getting into the the specific operators, you know obviously they provide a number of different functions for Postgres or MySQL and the various other ones.
00:44:30
Speaker
um Could you talk a little bit about sort of what functions those operators help with um in terms of like, you know, from day zero to day two, things like backup, restore, and other things?
00:44:41
Speaker
Yeah, ah these operators do almost the same thing, but for different databases. In the case of Percona ExtraDB Cluster, we have PXC, which we call it PXC operator. for The PXC operator simplifies the complexity of running MySQL,
00:44:56
Speaker
in a containerized environment, ah providing tools like backup, restore, and restore management. management In the case of a PSMDB, we call Bergona Server for MongoDB operator, and it managed replica set, scaling, sharding,
00:45:11
Speaker
and Oh, wow. Yes. It does configuration in Kubernetes cluster. It also supports high availability and ensures that MongoDB cluster is always a stable in Kubernetes. and The last ah is Percona PostgreSQL operator or Pg.
00:45:29
Speaker
Automates critical database tasks like provisioning, scaling, upgrading PostgreSQL instances, and enable also high availability as the other operators.
00:45:40
Speaker
yeah um in that that application. Interesting. Wow. ah It definitely does a lot of things. So ah two questions. ah One is oh if I already had databases deployed through the Percon operators outside of Everest, can I then start managing those using Everest? like Can this work in a brownfield deployment?
00:46:00
Speaker
Or do I need to start fresh and only use Everest-based databases? That is good question. I think this is a where we are not a jet. I mean, this is what we plan to do, that everyone can use Percon Everest and link their databases or their operators.
00:46:17
Speaker
So this is the goal for us. ah But for now, we are yeah supporting our role the Kubernetes operators. And can I deploy Everest anywhere, any Kubernetes distribution, or is it very specific to like, oh, I can only run this on EKS, for example? No, you can use Google, Azure. sure you can use a a i think, OpenShift.
00:46:41
Speaker
Now we are in testing OpenShift. You can run, vi already tested. But you know, way ah for production environments, I'm not sure it's ready. But for production environments, Google, Amazon, Azure, they are already Okay.
00:46:55
Speaker
okay yeah And I want to go back to like the question that Ryan had or are the task that Ryan had mentioned, right? Like the schema migration or ah database migrations. Can Everest help me with some of those as well? You mentioned sharding in case of MongoDB, but...
00:47:09
Speaker
What about some of the other database admin specific operations? Yeah, with Percona Average, you do it ah through the UI. You can do it in CLI as well.
00:47:20
Speaker
But in the UI, for example, you can create backups. You can create point-intel recovery backups. You can restore a from an existing database, from existing backup.
00:47:32
Speaker
You can also... and configure the amount of nodes that your cluster is going to have in Kubernetes. in the case of In the case of PostgreSQL, you can configure a PG boundaries. You can configure sharding. My MongoDB database is going to be sharding. okay how ah what How many config servers confi server my MongoDB is going to have.
00:47:56
Speaker
You also choose that. So you have just to go and choose and choose, a start to choose. And also you can add a PMM. You can add Percona Monitoring and Management. So everything is integrated. So you're just clicking activated Percona Monitoring and Management. But if you have already installed it, Percona Monitoring and Management, you just link it and put the endpoint into Percon Everest and it's already done.
00:48:18
Speaker
Also, you can give external access to people to your cluster so people can access. There is one one and more interesting thing that it it gives you. ah The details that Percon Everest gives you is a lot. So you can see there how many resources are using and also how you can connect ah how you can connect to your database.
00:48:38
Speaker
So you can just copy the and the the URL, go to the CLI, and ah start to insert data into MongoDB, doing all that. Okay. Okay. one One more question and I promise i'll i'll I'll go and read the documentation or go read the manual.
00:48:55
Speaker
Like, okay, we spoke about database admin specific operations, but does Perkona Everest also help me with Kubernetes cluster auto scaling? Like, obviously I can expand or scale out my database, but how does it handle the underlying cluster? Like, can does it work with things like AWS Carpenter or but something else that can allow me to scale my Kubernetes cluster if I need more ah more compute or more storage, I guess?
00:49:18
Speaker
ah You have the option to to change the size of your cluster in Percona Everest. You can change, you can add more CPUs, more memory, it depends depends on you. and The amount of nodes also so you can change, ah you can even customize them.
00:49:32
Speaker
a But I'm not sure if you can ah scale, like you say, in in another database, ah if I understand correctly your question. Scale the underlying Kubernetes cluster?
00:49:43
Speaker
like if If my yeah database cluster is growing, right yeah what if I run out of capacity on the Kubernetes cluster? Can Everest somehow automate that as well? ah i I don't have the answer for that. okay I think it's time to open up here, Ryan, against the Orkona project. yeah i have I have to check if that's Salah also.
00:50:04
Speaker
Thank you for the for team pointing out. yeah you but Yeah, those are those are all like really interesting use cases. and And I think this is where, to your point, Edie, before, is that like you know you're you're you're talking about all these features that um Everest does for individual databases, and there's ah there's a laundry list, right? There's a lot of things.
00:50:24
Speaker
um But you know ah when you're operating it day-to-day, you know're you're going to run into various things like you know inserting data or even doing migrations and As someone who's using Postgres for an application and and using things like Alembic to do migrations for Python-specific implementation, I don't think those things are going to necessarily... i mean maybe they Maybe they do make it into an operator. I'm not sure, but um it could be application-specific. That's where like the intersection between like the the operator and then the DBA and then the application team, all working on Kubernetes and maybe a platform engineering team, kind of they all...
00:51:03
Speaker
have a role still, right? And they I think there has to be understanding of who takes care of what. And maybe that's more of like a more like knowledge-based building, you know, through a community type of type of exercise. i'm not sure.
00:51:19
Speaker
Yeah, that that is how how it works. Everyone i play a role into this ah and implementation of Kubernetes. Makes sense. So I do have, um you know, I mentioned that ah you are a DOK ambassador, and I wouldn't i wouldn't be doing my job if i if I didn't ask you, how does your role as a DOK ambassador sort of ah play into your work with Berkona or even the open source community in general?
Standardizing Kubernetes Operators
00:51:45
Speaker
Yes, you are ambassador as well. Yeah, glad to see you there in the community. I'm ambassador for the OK community. As a community, we are working together, you know, in find ways to collaborate, to create paper research, to create content, to so have meetups, interview people and trying to connect members from our community.
00:52:09
Speaker
to create working group together and with people who are interested in S-Rex, big data and applications of data, also in data on Kubernetes, whole workflows run in Kubernetes.
00:52:22
Speaker
For example, one of the things that that we did together as a community was a very very nice ah timeline ah for where we put together all the Kubernetes operators for databases.
00:52:36
Speaker
So we put it together. So this was a community work and we pushed it into the CNCA blog post and it's there for celebrating the 10 years of Kubernetes.
00:52:47
Speaker
Yeah, god's love we've certainly come a long way, that's for sure. Yeah. and Another thing that I also saw in the community is that we're working together and finding these spaces to talk about the standards.
00:52:59
Speaker
and For example, and what we what we want to achieve is as you said you as you said, there are several initiatives to create Kubernetes operators for different things.
00:53:11
Speaker
But ah we don't find one standard that says, hey, this is how we should we shall do it and we start to implement that way. Everyone is running in his own way, which is also amazing because there are different perspectives.
00:53:24
Speaker
But this is where we are working now as a data in Kubernetes, to try to find these spaces to discuss and also to find how we can pass through the next step to standardize the process to to build Kubernetes operators.
00:53:39
Speaker
Yeah, that's ah that's a fair point, right? The operator the operator pattern that um you know goes into developing operators gives you a lot of flexibility. And just because there's one operator for Postgres and another one for Postgres or another one for the... They're all not built the same and off the same necessarily standard to provide the same type of things. And so like ah maybe that's more of a...
00:54:02
Speaker
ah a you know that puts the onus on the person doing the research, right. Of what operator they want to use um really depends on what they're looking to get out of it and maybe where they see that thing going. Cause there is, there is some sort of, I won't, I won't say, um you know,
00:54:21
Speaker
you have to keep using one operator, but if you start using it, you're certainly going to get familiar with it and you might as well, yeah you know, make sure you're, you're, you're using one that's being developed and kind of continue to be supported. So.
00:54:33
Speaker
Yeah. later has ah He's always strange, strange. And, ah you know, it is gonna, the, the work that you are going to do is gonna depend of how well the operator is righted.
00:54:45
Speaker
Yeah, yeah that's that's a good point. So any of our listeners who are either looking to get in the the data on Kubernetes or databases on Kubernetes um you know community, come talk to Edie, come join the DOK community. um ah you know You'll find the three of us there.
00:55:04
Speaker
um Even if you're sort of on the DBA side or on the application side, I would also say come join all right and ask these kinds of questions that you might have that you you know didn't get out of this ah podcast or even out of your the communities you're part of. but I think we can all say that we'd welcome you there and and love to have you there.
00:55:22
Speaker
um So, Edie, where where can people learn more, um either to find you, more about the Percona project, more about anything else that
Engaging with Percona and Community Resources
00:55:31
Speaker
you want? well Anything you say here, we'll put in the show notes so people can find it.
00:55:34
Speaker
Yeah, I will share the links with you. So you can visit easily our documentation. But I suggest you to check our Percona Community Forum, which is which has a huge ah community members there so people can discuss about our different projects, all our projects in open source.
00:55:52
Speaker
So also GitHub repository is a good way. People like sometimes interact with GitHub and start to raise issues and conversations, right? I like also our Percona website. You can visit percona.com.
00:56:04
Speaker
And we are in all our social media. Now we open Percona account in Blue Sky. in the sky. ah wow yeah I hope you are there. I will start to follow as well if your friends might have the sky.
00:56:19
Speaker
And um for me, yeah you can just type my name in LinkedIn or when you can find me. right Great, great. And of course, if you're at various KubeCons, I'm sure you can find find you there as well. Yeah.
00:56:33
Speaker
In KubeCon, in Londres, everyone is invited to be there. So it's going to be really amazing topics. I'm going to be in a Clone Native University co-located events as a speaker.
00:56:46
Speaker
So I will talk with Deepja and with other members of the community about how we activate community ah to grow in Clone Native. It's going to be a amazing panel discussion. and um Also, you will come to the Dayton Kubernetes is Community Day as a co-located event as well. It's going to be the first ah of April.
00:57:05
Speaker
And oh this is going to be amazing because a lot of people is going to talk about workflows or Dayton Kubernetes. So you are invited. I don't know, man. Ryan or I won't be there, so I don't know how how great it will be.
00:57:18
Speaker
I'm just kidding. Oh, you you'll be there. So it'll be just as good. I'm lucky because I live here. ah That's true. Yeah. I'm just kidding. Like DOK days are always great. right like They have a ton of great content. I always like, even if I can't attend in person, I always watch the YouTube videos. Yes. too. same. very useful. to Well, ah Edie, thank you. I'm glad we finally made this happen for 2025.
00:57:46
Speaker
and we'll we'll definitely have you back on the show at some point. But thank you for coming on the show. and we appreciate you. Yeah, thank you so much. And have an amazing day, Bahim. I hope this is not the first time. I hope we have another podcast.
00:57:59
Speaker
But yeah, thank you so much for the invitation. course.
Percona's Open-Source Commitment
00:58:04
Speaker
but All right, Bhavan, I knew I'd enjoy that conversation with Edie, but what were your takeaways from that? I think I like ah the Everest solution that that Percona has oh and and the commitment behind open service, right? So obviously,
00:58:20
Speaker
oh having multi-operator offering. ah Usually when we see vendors do that, they usually close source it and have to and have ah something available for free.
00:58:32
Speaker
I think when Edie was talking about the commitment that Percona has ah behind open source, I think it makes sense that, yeah, they want to do everything in the open in open, have like a...
00:58:44
Speaker
graphical user interface, have Kubernetes-based resources that they can use. I'm sure like they'll make money in in terms of support and services, but still, like being part of this cloud-native ecosystem and pushing for data on Kubernetes in general, being a major contributor of the Valkyee project, I think all of those things ah definitely, like I don't know, ah back the fact up that, yeah, Forkona is invested in this ecosystem for sure.
00:59:07
Speaker
Yeah, and definitely one of those companies that is well-renowned in in the community. So they're doing a great job. I think my main takeaway from that conversation was, you know, Edie is one of those people that interacts with the community quite often. and And to hear her talk about how her feedback from, you know, folks in the community are are still finding that um that gap in, I guess, knowledge or skill set you know,
00:59:37
Speaker
of sort of running databases beyond sort of the basics,
Kubernetes Community Growth
00:59:43
Speaker
right? um You know, as, as ah you know, Kubernetes is now over 10 years old, we've been having, you know, stateful applications on'm on on containers for even longer.
00:59:55
Speaker
um I think that just shows still how ah new the concept of a database in a container is for a team, not an individual, right? um I think especially you and I, and and maybe some of this community are are aware of of how things work, but hey, running this thing in production, um you know, operators get you pretty far, but there's still a lot of knowledge that, you know, a DBA would have that maybe they don't necessarily,
01:00:26
Speaker
fully understand the operator piece, but the the people who do understand the Kubernetes and operator piece don't have all the knowledge of the EBA. So it really just kind of talks about, um I think there's so much more than just um understanding the community space, but you know really having the deep knowledge and teams and expertise and having the DBAs come with us.
01:00:47
Speaker
We want them to be fans of what we're doing in the community space. And we want the knowledge um you know that they've had for running these things at scale for Many, many years, I'm sure. So um yeah, I think, I guess it's not surprising, but um it's validating to hear that maybe we're having more of those conversations. Like, hey, yeah, we're doing this.
01:01:09
Speaker
Now let's figure out how to do it really well at scale. no No, I like your point, right? The community is expanding itself and bringing in DBAs for the ride, I guess. Till now, as you said, it was individuals inside teams that were experimenting with Kubernetes and maybe doing some little work, maybe having new applications to broaden Kubernetes. Now with this expansion, I think we can expect more and more usage. So definitely a good sign.
01:01:32
Speaker
And we are in the right space. Great. Cool. So um for the listeners, we'll make sure and put all the links like ah the Valky project, PMM we talked about, Everest we talked about, all the Percona links that ED is going to give us and we'll put them in the show notes if you're interested in finding out any more.
01:01:49
Speaker
As always, please subscribe ah to the podcast and on YouTube. Thanks. We actually are um over 500 after one year, which I feel like is, you know, respectable for this show.
01:02:04
Speaker
We'll see. And as always, if you have any um ah topics you'd like us to cover for 2025 at some point, um please um just let us know on either website, Slack, email, whatever it is.
01:02:20
Speaker
And we much appreciate that. And as always, thanks for listening. um And but that brings us to the end of today's episode. I'm Ryan. I'm Bobbin. And thanks for joining another episode of Kubernetes Bytes.
01:02:35
Speaker
Thank you for listening to the Kubernetes Bytes podcast.