24:01 Video
Cloud 2.0: Data Strategies to Supercharge Your Cloud
The enterprise shift to hybrid cloud is real. But is your hybrid cloud environment drowned by silos and complexity? Learn how to streamline your hybrid cloud.
Click to View Transcript
Andrew Miller: Welcome to Pure Accelerate Digital 2021 and specifically to the session Cloud 2.0: Data Strategies to Supercharge Your Cloud. That's kind of a fun title, I think she'll be Oh, my name is Andrew Miller, Principal technology strategist here at Pure, perhaps most importantly, started on the
customer side for about seven years and engineered architects partner pre sales, I see manager director building out a tech marketing team. So I've been doing things that orbit around this topic for a while trying to put some stories as we go along the way. But I really enjoyed doing this session with Shilpi
just this is what you focused on in different ways for the last couple years. Shilpi Srivastava: That's right. Hey, everybody, I'm Shilpi and I lead Cloud Solutions Marketing here at Pure Storage, been here for three years. And it's been more exciting than anything I've
ever done before. first project here was, you know, when we launched the first container storage as a service solution, and now we've evolved into a full Cloud Data Services portfolio. So super excited to talk to you today. So let's just get started, you know, you're in the business or you know that,
for businesses, today, it is more important than ever to move super fast. And we know that a lot of enterprises now are driving towards hybrid cloud, almost everyone has workloads running in multiple private and public cloud environments. With that technology that makes applications and data mobility
simple, is playing a more important role than it ever did. Andrew Miller: There is sometimes terms are a little mushy, I know if I go back 10, 15 years, sometimes the CIO came and said, You have to have a cloud strategy. So we all did VMware. And then we kind of check the box that we had a
private cloud. So while the terms can be mushy, there's an increasing amount of definition here and even reality. In cloud strategy, cloud native, you go to sites like 12factor.net, talking about modern applications. There's much more reality here today. And that's what we're hearing from our
customers. There's still challenges, though. Shilpi Srivastava: So hybrid cloud sounds great, right? You can, in theory, you can get the best of both worlds, you can get that agility and flexibility of public cloud environments, while maintaining the security and reliability that you get for
your workloads on premises in your data center. However, when we look at a lot of these enterprise environments, we find that multi cloud is really creeping in a lot of complexity and silos. Because there's different application stacks in different environments, on prem, in your hosted cloud, or in
public cloud. There's different management and licensing. And the biggest problem arises when you're very valuable data is fragmented and siloed, across multiple environments, and diverse backends. Andrew Miller: I mean, we know that historically, we've had
silos and silos aren't inherently bad, right? If we have no silos, application, performance becomes a real challenge. But we don't want to bring the same unintentional choices and consequences into a cloud architecture necessarily, where it turns into a different version of you get a silo and
you get a silo and you you get a silo to you, that's not what we want to be unintentionally. Shilpi Srivastava: That's Andrews best impression of Oprah. So if you're, you know, we've just over the last few years worked really hard to help you unified your multi cloud
architecture, by using data. Because if you've been following us, if you've been to prior accelerate, events, you know that data is what we do best. And so we've been working to build a unified data plane, across your hybrid or multi cloud environment. And this data plane, the key attributes I
should have is, firstly, you've got to have storage as a service delivery on demand for all your applications traditional or cloud native, you've got to have a unified management platform, as well as unified licensing for your hybrid cloud. It's got to be simple when it comes to moving applications between on
prem and public cloud environments. So that you can unlock really a bunch of use cases, which we will talk about. And finally, you want to make sure that you can modernize applications, wherever you're running your workloads and build that cloud native enterprise.
Andrew Miller: This is the Holy Grail. In a perfect world. We've got a data plane across all the clouds of different kinds and types, how to find them. Easy in theory, easy to conceptualize, but super hard in practice, which is why as we're going through this today, we wanted to focus not just on the
technology, but actually have a use case focus to hopefully make it a little bit more real for you. Pretty shopping. Shilpi Srivastava: Yeah, so the goal, the final goal is really to unlock seamless mobility of applications. So that whether you have a scenario where you maybe want to develop
applications in the public cloud, but then deploy them on premises, that's super easy, right? Also, if you want to leverage the public cloud as a secondary or tertiary data center for your backups, or as aid, or for disaster recovery, we want to enable you to do that in a efficient as well as an
effortless manner. Andrew Miller: So when we for the rest of this session, this is basically your agenda of what we're going to go through, I want to test you know, cloud agility on prem. Because, hey, that builds on Pure history, getting some of those same
capabilities in public cloud. But what about if you're starting from a cloud perspective, cloud native, simplifying mobility back and forth, because that is the ultimate outcome of the previous three items. And if we started with the promises about cloud mobility, you might not believe
us. Now all of this is built on top of peers portfolio, their products here, we're intentionally not going to talk about them very much. Today, we're going to talk about use cases, integrations that go in various application and infrastructure and cloud levels. But of course, they're slash
play, there's flasharray, there's cloud block, store, Flash, racy, other things that undergird the capabilities that we're going to talk to you about today. So first section, cloud agility, but from an on premises standpoint, Shilpi Srivastava: you're we're, we strongly believe that the
cloud is not a location, it's a style of computing, right. And so what becomes super important now is to make sure that your data center is modern, and it acts like your private cloud. And therefore, and you can do this by building cloud, like agility on prem. How have we been doing this at Pure?
Firstly, we want to make sure that whether you're consuming services on premises or products on premises or in the cloud, you should have a similar experience to kind of unify those environments. And so we've built an AI assisted service defined digital marketplace through Pure1, which is also integrated
with Pure as-a-Service. So which is our unified hybrid cloud subscription service. This is just like an AWS or Azure Marketplace, but it's kind of elevates that to the next level. Because here, you can just have a singular place to buy anything you want. Any infrastructure you want to run, on premises or in
the public cloud environment. We also want to make sure that we completely integrated with VMware vSphere, which a lot of our customers used, and with Kubernetes, orchestration platforms across the board. And finally, we are integrated with infrastructure at ko escort platforms and tools. And you can
see a bunch of them here on the screen one going to each of those since we're really short of time here. Andrew Miller: That's really the ones that you see there are our because of here's API first approach from an automation standpoint, that API first approach is, to me part of what
makes pure on prem, not a product, but a platform. You can integrate it with anything. But also more importantly, as you now see here, it's as much of storage that acts like software as a service as possible, can deploy it once upgraded and place it gets better over time, without taking you know the
technology taxation of planned downtime or forklift upgrades, data migrations, when you want to embrace new technology, even during those migrations, there's not performance impact, not just no downtime, but no performance impact, as well as that as it improves over time. That means you can drive much longer
depreciation schedules, if you're in a capex construct. That's that multi Gen life's generational lifespan investments protected. There's hard architectural work done that sometimes I'll do 3060 minute two hour deep dives on, we won't do that today. But there's real architecture work
that has gone into really the last 10 years of why pier has been successful. But you mentioned that the service defined marketplace shall be pushed even further than kind of the on prem platform. So in that service defined marketplace, you know, there are a lot of pieces there and Pure1 from an
optimization standpoint, streamlining, empowering, right kind of thing. This gives us visibility across the stack and various levels of automation. Shilpi Srivastava: And the key here is that to note here is that we're really bringing the cloud experience on premises.
And the next slide is what it actually looks like. So this is what the digital marketplaces so you can see you go in there you choose whether you're looking for a Kubernetes backup, but are you looking for Kubernetes disaster recovery, block storage as a service. And the coolest thing is that it's all self
service. You go in there, you click what you need, and you just get it. Also, it's we're now completely integrated with the AWS as well as Azure marketplaces. To give you that singular hybrid cloud view. Andrew Miller: I feel like that's the perfect segue to
section two which is Get more from public cloud, and how we've been braced public cloud in the cloud, not just from a marketplace standpoint. Shilpi Srivastava: But exactly and we were talking about, when we started off, we were talking about a unified data plane that
only works if your on prem environments has that cloud like agility that we spoke about. But then also, you've got to make sure that your public cloud infrastructure is on par with the infrastructure capabilities that you expect for applications running on premises. So with that in mind, about a year and a
half ago, we launched cloud block store, cloud block store is Pure software running in a public cloud environment be that AWS or Azure cloud, and I just designed to help you get more from your public cloud storage investment, we're really making public cloud better. Because cloud block store is designed
for high reliability as well as as well as predictable performance. We have the dual controller architecture, which is very robust and highly differentiated, we'll get into that in a bit. Also, we're bringing purities leading data efficiencies to the public cloud. So now you have thin
provisioning, deduplication, and compression in the public cloud, in addition to our spacesaving instance snapshots, which are super handy when you're looking at creating dev environments, while also reducing total cost of ownership. And finally, we brought in Pure HA capabilities, as well as software based
encryption capabilities. So that your public cloud environment now starts looking a lot like your has all the enterprise capabilities as your on prem environments. And when we look at the architecture, for cloud block store, we made sure that we are leveraging the best of two public cloud resources to
make that happen. Andrew, can you take us a little bit into the architecture on Azure and AWS, Andrew Miller: happy to I love what you mentioned about that our goal is to help make cloud storage better to add capabilities that reflect what we've been doing for years.
That's why it is an Azure, they're bigger, this new here, they partner with it. So when we look at the architecture, you actually think about how would you embrace the cloud from the perspective of what the cloud is not just try and pick it up and run as a virtual appliance. But embrace the cloud is a
collection of virtual hardware primitives. That's not a derogatory term. That's actually what AWS and Azure say we give you primitive services, please build cool things with them. So we did this. But if you look at that, you know, it's a dual controller architecture similar to on prem, we've got virtual
shelves and virtual drives. From an AWS standpoint, we're actually consciously understanding both EC2, EBS instance, storage, the different types of EBS, I01 versus GP2. So we can actually balance performance scaling and capacity versus cost. And then also even leveraging S3, like you say, you
know, there's 11 nines of durability, we persist data in S3, long term to make sure we never lose your data. That's a critical thing when you're when you're a storage company, right kind of thing. From an Azure standpoint, there's some similarities. But there's also some differences. to still do
controllers, we have virtual shelves, we emulate the same Nv ram capabilities, which is how we handle inbound rights to make sure that once we've acknowledged a right that we never lose it. But then also some unique pieces there you see Azure Cosmos DB listed and Key Vault, even some pieces around
just in time, access out of certain levels of Azure AD for remote access, in order to be able to get into your environment only when you would give us permission to do so. Now, we talked about the benefits, we talked about the underlying pieces that are different. And sometimes I like
to whiteboard this part out really long. But there has to be as well be a financial piece to this. So the technical flexibility, we've talked about a good bit. But that needs to be paired up with financial flexibility, which we deliver via Pure as-a-Service you see there. This is a consumption
based op x model. That is gives you the same technology between on prem and the cloud. But it gives you a unified cloud model by virtue of even having a unified subscription, one subscription between on prem and the cloud. You can even rebalance as needed. A lot of other capabilities that you see
listed there. But that's the financial complement to the technical flexibility. And you can even leverage this on prem, even before you start to move your data into the cloud. Now, we did tease a little bit of this earlier, and I think we already said Kubernetes. So this can be the second mark on the
bingo Kubernetes bingo card for anybody. But you've even done some whiteboards on this one too, I think right? Shilpi Srivastava: Yes, that was so long ago. So over the last few years, we've all seen containers and Kubernetes going from that early adoption stage to becoming mainstream at Pure
we caught this trend early So back in, I think around four years ago, we released our first Docker volume plugin, and then flex volume drivers. Remember that. And then Pure Service Orchestrator was the next evolution, which allowed you to automate storage provisioning across a fleet of blog and file
areas. But we didn't want to stop right there, we want to take this to the next level, we wanted to provide a complete set of capabilities, data services, and data management on the application level. And so we're going to started looking into how we can provide this Portworx became an obvious choice. A lot
of Portworx, customers were showing their preference towards using Pure as the back end. And so the Portworx was a great fit, because just like everything else at Pure, it was built for business critical or tier one applications. So it helps developers build and update applications faster, without
having to compromise on security, and enterprise policies, as well as it gives your SAS architects is scalable, and reliable Kubernetes storage platform. Andrew Miller: Also a little bit of a fun fact, this is one of the few slides, you'll see. If you look in the bottom right
there that has those logos on it, you don't usually see those on a Pure slide. that's intentional, though, because we actually want to be able to help and Portworx was before we acquire them. But we want to be able to help customers that are on other storage platforms that maybe haven't chosen Pure
Storage for that layer yet. But but like you mentioned, she'll be actually I believe it was the biggest Portworx customer chose Portworx and then independently chose Pure Storage as a storage layer because there was so much alignment there. If we go into how this works, there's a little bit of kind of an any any any
flexibility message here, any application, any Kubernetes, distro, any on any cloud, any infrastructure, the different stages of development cycles. And then these are broken down into models inside modules inside Portworx. So initially, a PX store you see there is focused on the initial
provisioning and container storage, then we think about protecting the data. So that's around PX backup, that's on PX DR. Even with PX backup, there's some neat pieces that go further up the application stack from an awareness standpoint, DR. You know, that gets between availability zones, and possibly
even regions kind of thing to zero RPO. Next, what about data portability, this could be between clusters, this might even be between clouds, you got to think about the egress charges, right? But there may be scenarios where you need that. And we can facilitate that. Security is still important,
both from an encryption, security policy and multi tenancy perspective. And last but not least, is PX autopilot. This goes into economic efficiencies about actually scaling auto scaling storage as your app grows, that helps reduce costs. And there's actually been times where we've
had customers, where they've actually paid for Portworx by virtue of the money they've saved by that mapping more dynamically, to what they need provisioned, being on demand versus being provisioned ahead of time. Man, I feel like we went through a lot, they're really fashioned up. But we want
to make sure to get to the last section, which is effortless data mobility. So we made some big promises at the beginning. We know we want a data plane, we know we want to be able to move things around more easily. And that brings us all the way back to the use cases that I think you talked about Shelby,
Shilpi Srivastava: yeah, and that this was what he pitched as the ultimate goal right behind building all of this technology, the cool technology that we just spoke about. One of the things that was super important for us, was to make sure enterprises were able to now run tier one applications in their AWS or
Azure cloud environments, and do that with ease. And with running enterprise applications, there comes the need to enable a che within the virtual array as well as across availability zones. So why did we do that we've got something super cool, called active cluster, which you might be familiar with if you've been
following Pure. This is active active replication, which you can enable between availability zones in the public cloud just as we enable it between areas on premises. Portworx also has also provides active active synchronous replication for your Kubernetes based applications across multiple public clouds,
or on prem infrastructure backends. Data Protection is that it's really top of mind these days, and by leveraging public cloud for Dr. And backups, businesses are able to save themselves, the resources and the cost of maintaining that secondary data center, Cornerstone one of our
customers, a leveraged cloud block store to deploy a hybrid cloud disaster recovery setup using cloud block storage In the cloud and having flasharray on premises there, they were able to really simplify replication between these two environments. And that really helped accelerated their journey to
hybrid Dr. When you look at, if you're running Kubernetes applications, Portworx provides complete data protection services for your applications running on Kubernetes. As well as all of the data that goes with it. Royal Bank of Canada was faced with the challenge of meeting very strict availability
requirements with just Kubernetes. So these are using Portworx DR. So an RP SDR. And with that, they were actually able to achieve zero RPO and an RTO of less than two minutes. So that means that they are now able to recover their applications in another data center, under two minutes with
zero data loss, which is super critical. And what this huge win for them. Andrew Miller: Think about running tier one applications, tier one application requirements around data loss and coming back online, that doesn't change. So that's
amazing. Continuing on, so we think about what about I want to develop in one spot and I want to deploy somewhere else, I need flexibility. So let's say I've got applications that started on prem. And then I want to deploy them or even or even migrate them to the public cloud, think about them in a hybrid cloud
standpoint. This is where actually peers asynchronous replication you see listed there has has leading data efficiencies from a replication standpoint, the amount of data that you have to put across now newskin is listed there. fascinating story in that actually cloud blocks,
replicating from on prem flasharray to cloud block store in the cloud, actually helped them both with their migration into the cloud, the efficiency of the replication. And then once they were in the cloud, gave them capabilities that they'd had on prem, that if they hadn't had in the cloud, being
able to clone certain databases 50 6070 times. That's why there's that 80 to one data reduction, you see there, that's where that comes from. Without those capabilities from a cost standpoint, and from a performance standpoint, they would not have been able to move into public cloud for certain
applications. But what about the other way? What about what if we're thinking about this from a from a perspective of going the opposite way of I'm developing the cloud, or I'm doing next gen application development, that's all you know, microservices, and Kubernetes, and containers, that kind of thing. So you know, I'm
actually maybe developing on Azure, AWS, Google, but I actually want to run in a different cloud or different availability zone or back even on prem, that we can give you that same portability that same you know, substrate that data plane, we promised that the beginning right kind of thing
that you can actually be able to pull off that experience. Now, hopefully, we're right at the end here. We started out here, talking about, we believe that we can help you develop a deliver a cloud like experience everywhere based on on premises cloud agility, making the on prem, more cloud, like getting
more from your public cloud storage, offering similar similar capabilities you've had on prem, if we embrace cloud native apps as being first class citizens, how do we fold that in a first class way as well as then going and thinking about hybrid cloud mobility? That's kind of the capstone of all the
use cases that we wandered through. I think we've covered more than enough Shelby, I'm going to turn it back to you to bring us home. Shilpi Srivastava: Yep. Thank you, everyone for joining. This was a great in a short time, we just wanted to introduce you to
a bunch of cool tech that we've been working on. We have a bunch of other sessions, demos, and documents on purestorage.com/hybrid-cloud, that website is on your screen. Also, if you have any questions for us, feel free to reach out via Twitter. You have our Twitter handles right there and
enjoy accelerate. Thanks, everyone.
A modern cloud strategy needs a consistent experience across a hybrid or multi cloud deployment. In this Accelerate 2021 strategy session Shilpi Srivastava, Director, Product Marketing and Andrew Miller, Principal Technology Strategist at Pure Storage, will discuss strategies to deliver a seamless cloud experience in your data centre.
Learn how to build application mobility into your cloud strategy from the start, and take the complexity out of managing heterogeneous cloud environments to future-proof your cloud journey.
Meet with an Expert
Let’s talk. Book a 1:1 meeting with one of our experts to discuss your specific needs.
Questions, Comments?
Have a question or comment about Pure products or certifications? We’re here to help.
Schedule a Demo
Schedule a live demo and see for yourself how Pure can help transform your data into powerful outcomes.
Call Sales: 800-976-6494
Media: pr@purestorage.com
Pure Storage, Inc.
2555 Augustine Dr.
Santa Clara, CA 95054
800-379-7873 (general info)