00:00
Welcome again everyone. Um we live in a competitive world. The fast are eating the slow and the fast and smart companies are racing ahead of everybody else. Going fast with technology today is about automation using technologies like containers and kubernetes and the smart companies are layering on ai and Ml to win with their
00:27
data. So I'd like to invite you sometime today to make your way to the rear end of this tent, to the portrait's pavilion in the back there, you'll see how portraits can help you with containers and kubernetes to deliver data management, data protection and data services for modern apps. Get them to production and scale quickly with Portrait Works.
00:58
Think of Porter works as your data partner for containers and kubernetes. So what I'd like to do next is showcase how Comcast went smart and fast for their video on demand business with the help of court works. I think if you look at the video, you'll see that I'm wearing the same shirt that I wore today, that when I was interviewing paul Davis, we focus at pure Sure,
01:27
love our orange. So here's paul Davis. Really excited to chat with you today, paul. Looking forward to this. Thank you for having me murli. I'm happy to be here paul most of the U. S. Audience, of course knows about Comcast, but for our global audience, can you tell us a bit more about the company?
01:48
Sure. Comcast is headquartered in philadelphia and it's the second largest broadcast in television cable television company, but it's also the largest media and entertainment company serving 53 million subscribers around the around the world contest is also the parent company. Too many names that you may have heard of before, like Sky Nbcuniversal,
02:12
DreamWorks and many more. Of course, Comcast has been a portal works customer for about five years plus now. And it's been a great partnership, you're a great example of what we call a co create partner and co create partner customer is one who has helped us test and develop right from the early days our solution and helped improve it and it's and it's been something that we've
02:37
done with you for a while now over the years. So while we kind of look at where we are today, can you, can you roll back a little bit and tell us how this whole thing got started? Sure. As you mentioned, we started working with four works about five years ago the catalyst to start working with networks was too,
02:56
I was put in charge of taking apps that support video infrastructure to a next generation infrastructure platform like Kubernetes. Well, evaluating options for kubernetes for persistent storage performance that works provided really led to that question, being asked, what else can we use it for? And that conversation led to a P O C and further building a cloud
03:24
based cloud, like excuse me as a service platform for observe, ability supporting the video infrastructure that I was part of. So how did the pOC workout really well, um that pOC turned into a full production project in one month. And the underpinnings of the platform is what really helped. So it helped Comcast save Melanie's and
03:51
operational um cost with the observe ability program the system provided the ability to quickly identify and identify issues that are long running as well as new issues that popped up. And that really helped to reduce that overall failure rate to under five under 1%. So a pretty big difference from where we started.
04:15
So the people who were affected by these failures were your video on demand customers. I assume these are the folks who are taking the time on Friday night to go watch a movie and really don't want to see a failure happening during that time. That's right, right. When you consider failure rate uh in percentages it doesn't seem very big. But um if you can imagine 1% of 53 million
04:40
subscribers playing something and um 1% of that failure is still a pretty big number. So bringing it down as close to zero as possible is very important. So those those affected were customers that are that are trying to watch contest provided content time is limited for all of us. So when consumers spend time to watch these, this Comcast provided content,
05:05
it's important that the content is available and ready to play when they when they're ready to watch it. So very mission critical and uh after that early success we realized that we could make a business case for how to evolve and expand into other business areas within Comcast. So now we uh support voice, we support networking, we support advertising, IOT apps and online apps
05:34
along with legacy video apps. Uh that that we started with. So paul, tell us a bit about your technology strategy and production. So the system is based on kubernetes and ford works but also elasticsearch in Kafka to name the core pieces for work really made it simple. It looks like local storage and it's easy to leverage.
05:59
So we don't require a separate storage team to manage and operate our networks clusters. We can move things around in kubernetes without impacting our service. And really that piece of scale that we get with court works has been instrumental in our success in 12 months. A key data point here is that we have more than doubled our utilization and I'll get them the numbers in a second.
06:23
But today we touch all lines of the business and we are located in the US and eu we're also expanding into e M E A and to Australia. So that's what's going on. That's a that's a big environment. So, you know, lots of demands I I presume on not just performance but reliability, availability and security.
06:45
Yes, it really is a big logging system. Um I guess to mention some of the numbers we support over 175 terabyte of in just per day and we store over 4.3 trillion documents. We in just greater than 2.5 million events per second. So it's a very complicated and large system all underpinned by works.
07:14
That's amazing. You know, I know from the perspective we aim to be a customer first organization and you should have helped us kind of build reliability, scale and and and availability into our product. So it's been a great partnership for us. Yeah, we are definitely a demanding customer for sure. We don't mind taking some risks to aggressively push our agenda forward and court works has
07:39
kept pace with us and embraced it. So that has definitely made it a great partnership. In fact, the new PX fast system is a direct result of this partnership and collaboration. So we push hard and put works has been there with us making improvements in performance and scale that advanced users of kubernetes will definitely be benefiting from.
08:01
So as mentioned, it has been a great partnership and it's been that way since the beginning. So we appreciate that paul. It's been a great partnership indeed. I really appreciate all that you've done for us. Thank you so much for sharing your story with us and and we really do appreciate it. Thank you. Thanks for having me.
08:24
Yeah, that sure is a great success story for paul. Davison. Comcast. Next I'd like to welcome to the stage, my colleague sir, be paul who runs product marketing for poor trucks. Welcome sir. B hey molly! Thank you so much.
08:54
Hi everyone. Glad to be here and excited to hear the partnership between Comcast and fourth works. So can you share some of the announcements that court works made recently and what are some of the key problems that we solve for customers of communities? Of course? Um it's truly exciting to see that part works.
09:14
Data services periods in short, fully supports six data services on one platform including cassandra, Postgres, Rabbit Mq, Casca readers and zookeeper. So if you're a devops team and you want to get empowered to run one platform operate one A P I and manage one solution for multiple databases going forward. PDS is the way to go and you're not going to stop here.
09:43
As many more databases and data services will be supported by PDS in the near future. So for all you develops and site reliability engineers the future is here. And as you saw in the demo from the keynote, Porter books, data services is so simple that anyone can do employ a new instance of a database within seconds with a few clicks. So get started immediately with data services,
10:14
one platform for any database and contact your poor work sales rep. Hey, fantastic. So congrats to the port works Data services team. PDS does allow fast and simple operation and deployment of databases now. PX enterprise. Our flagship product is something that allows customers to go from production really
10:40
fast to being scaling even faster. So what are some of the new use cases we support for that. Yes, really. So if you're an application architect or a developer or a platform owner who is building or wanting to build communities apps in production, we have announced five new use cases of court works enterprise including multi tenant
11:05
platform as a service, five G and edge lift and shift of VMS to containers. Uh, unified storage overlay for containers and next gen, next gen performance and scale apps. And this is all going to be supported by upcoming rich capabilities of features of Port works enterprise. Right, Really exciting. So as we see customers move to production,
11:31
one of the key concerns that everyone has is data protection. Now, kubernetes poses some unique challenges in this area. What do we have for customers here? Um, well for customers who need an easy button for protecting cuba needs apps, we are very proud to have recently announced general availability of support Works backup as
11:55
a service Bass for any Cuban every services. Um, it's based on the foundation of PX backup, which is the market leading Cuban eddies data protection platform. So portals Bass unlocks plus amazon Auto discovery of amazon EKS clusters today and one click backup and restore operations for all culinary services anywhere on prem or in the cloud of your choice And in the future we expect to support auto discovery
12:25
for other culinary services as well.
