Sathya Sankaran, Catalogic | KubeCon + CloudNativeCon NA 2022
Sathya Sankaran, of Catalogic, GM and CloudCasa, joins Alan Shimel at KubeCon to discuss CloudCasa and what it does to help solve the lack of backup options for Kubernetes and clusters through its backup-as-a-service plan.
Transcript
This is Textron TV. Hey everyone. We're back here live in Detroit at kubecon.
We're out on the floor. Luckily our Mike's doing pretty decent job of filtering because there's a people mover that goes up. It's about three hours but it makes a lot of noise.
You're not hearing it at home people here here it though. They got a little freaked out and just the den of people on the floor here, you know, it's right after lunchtime a lot of people heading back to sessions, but I think they said there was about 7,000 7,000 on site and about 11,000 or 12,000 words registered virtual. Yeah.
So it's a crowd. It's a lot going on. I'm happy to be joined by Sofia Sankara.
Sofia is a CEO a cloud casa. The last time something and I spoke was in Valencia a cloud a kubecon Europe and unfortunately. I don't know.
We may see each other. Maybe an answer. Yeah, but anyway sassy, you're welcome back.
I hope you did. Well, thank you. So.
We need to start off by reminding the audience about Cloud concept. Yeah, right. I know the story but they may not so why don't we start there?
Yeah, I mean look closer is essentially a secure home for your backups in the cloud. Right and and all of these homes need to be both secure and smart these days. Yeah, and you're an operating in that kind of an ecosystem, right?
So what cloud cloud costs are delivers is backup as a service. Okay, there are several Solutions out here that are self-managed. But you know in In This Cloud native ecosystem everybody is going Cloud first according to the last cncf service 79% of the respondent said they rely on managed kubernetes or hosted kubernetes.
That means what presents 79 wow. So that's a huge percentage of users that are putting data in the cloud first and if something isn't solving their problem, of course, you have the option of running it on Prem but the data is being born in the cloud with modern applications. So what clubs are does is Leverage is the same model for backups.
We are Cloud first we operate in the cloud we are With kubernetes for kubernetes. We are kubernetes dedicated backup service. We provide backups for both your on-prem workloads as well as multiple Cloud kubernetes engines The Big Three AKs eks gke we provide backup service for you and we provide hybrid cloud and multi-cloud Mobility which allows you restore any of these workloads in any of these platforms and and that's what cloud causes us.
Excellent. You know, some people watching this they say back up. It's been around forever.
Not sexy. They solved that problem a long time ago. Not really.
Those are people who have never had a rely on a backup. I mean living in Florida where we have hurricanes. I've had the unfortunate.
Experience. Yeah having a resort. Yeah.
Well, no recently. Didn't hit us. Thank God, but this is going back to 2004 2005 been a bad season with hurricanes those years and what you come to find out is though you paid for back up.
They didn't have back up his service then you had tapes or this. But it wasn't it wasn't an easy thing to restore. Yeah.
Yeah, right and I was gonna say thanks with kubernetes. We've actually gone a step back in terms of recovery options. Yeah with physical servers and VMS.
We kind of like this is where the approach of hay backup is boring comes from because we've solved a lot of the crucial problems right with VMware. We solved agent. Let's backups.
Yeah, he sold instant virtualization. Yeah, we provided bare metal recovery. All of these all problems were solved it with physical servers and VMware, but it's not true about kubernetes.
Look at how many solutions do cluster recovery. We do it but I don't know of any other solutions that do it how many people do cross-cloud cluster recovery again, we've we're finding some of these recovery capabilities, but it doesn't didn't exist until we brought it together. So the equivalent of a bare metal restore doesn't exist.
The equivalent of incident virtualization doesn't exist. So the rtos are still pretty long. It takes a long time to recover a kubernetes cluster today unless you have a stand by cluster sitting around but stand by clusters costume money in the cloud.
Yeah. They don't just sit around cheap. Yep, sit around idle.
I mean that I know that sit around I know if you stand it up someone will use it for something else exactly and and then you don't no longer have control of it and so in My thing about this this problem, you know standby nodes and Recovery nodes is that people are really really good at asking you for something when they need it, but when they no longer need it, nobody's really diligent about letting you know, just sit around and proliferates and and then you're being here every time paying for all of it in the best. So we are thinking composable recoveries, which means we back up your workloads kubernetes workloads. On-prem Cloud doesn't really matter.
But when you want to recover we want to start from point zero, we don't want to say have a functional cluster up and running have a standby closer to have a recovery cluster. We're saying we don't need anything in your infrastructure. Give us your Cloud account and we will build it for you based on how you're Source clusters.
We're sitting there love it. And that really is back up as a service exactly. Yeah.
What about you with the show this year? New announcements or yeah, big big announcements for us. We were always trying to optimize our service for the guys that are running kubernetes in the cloud.
We talked about in Valencia about adding Azure kubernetes service on top of Amazon's kubernetes services. So we did eks and AKs in the last yeah last time we spoke at Valencia, of course, the natural extension of that is delivering gke so we Now cover the victory platforms previously at Valencia we talked about hey how we do recoveries within the platform so I can take a cluster that is running in one account in AWS to another account in AWS because people do maintain a degree of separation for prod and non-prot environments. So you back up and restore to different accounts, but now given that we have direct integration with all three providers.
We're also providing a translation layer where you can take an eks cluster and restore to AKs we can take an AKs cluster restore to gke and why Sports And the last piece is we are also enabling Cloud migrations themselves, which means you can protect a kubernetes cluster on Prem. In future you say hey, this cloud provider has solved my problems. I won't actually move to him now.
We allow you to back up something on-prem and restore that to the cloud. So a lot of attention story and new record capabilities, but also To kind of filling out the dance card. Yeah, you know what?
I mean? If you doing it in AWS and Microsoft you gotta have any Google too? Yeah, right.
It's it's it's that maturation of a product right? Yeah at the end of the day, we're an insurance company, right? So you have to provide coverage for every possibility here for saying that they'll try to license.
Yeah, you know need more regulations. Hey, but I I did not mention the website. Yeah, so we are at Cloud cost at IO.
As loud Casa yellow udca essay Cloud outs dot IO. Yeah, and all you need is an email address to log in. Get an account and 15 minutes you could be doing backups and we will help you recover.
Absolutely. It's pleasure seeing you again. See you again.
I hope it's won't be Amsterdam that I see you next but if that's when it is, we'll see you there. It's not a bad place to meet you can always zoom in during the week to a text drug TV segment. All right, we're gonna take a break here at Detroit.
I think we have a little lunch coming up. We'll be back though, continuing our live coverage from kubecon.