13:32 Video
FlashStack: New Enterprise Applications and Future Plans
Watch this session to view an overview of the latest applications that have been tested and documented in Cisco Validated Designs covering VDI, Oracle, SQL Server and more.
Click to View Transcript
Mm Hi, I'm john McKay, Bill. I'm a senior product manager at Cisco and I manage a lot of the enterprise solutions that we sell on flashback, which is an integrated infrastructure that we're here to talk to you about today and really what I'm going to be covering the next 10 to 15 minutes and some of the latest
solutions we've done on flashback as well as giving you a hint to a couple of new solutions that we bring the market very soon. So let's get started with talking about Cisco validated designs and we're talking about validate designs. Were talking about extensive testing that has been done by Cisco engineers in Cisco labs, fully documented by Siskel.
The goal here is that the solutions that we do across our solutions on flash doc have been developed in these labs where all of the components have been listed in the document as well as all the version numbers of software and so forth. The idea is that you as a customer should be able to take this document and use it to implement flashback in your data center. You should get the same level of performance
that we've documented. If we don't, you call Cisco tax. We will work with you on the phone or if necessary. Even put an engineer at your place of business staying within your covid protocols of course. People to help you get the exact level performance. We've documented Cisco validated designs are something that the industry has really embraced
over the last decade and these are the documents you want to consider and read thoroughly as you look to implement these solutions for flashback. One of the key things about all the Cisco quality designs that we do with enterprise applications is that we want to document that there's not a single point of failure in a flashback design. And so we go through and we implement service
failures into the design. You know, we'll go through a design and we'll make several servers to fail simultaneously. We want to show that to UCS manager Francisco inter site. We're able to document the fail over that occurs when a server fails so that your workloads stay up and running you as the end user should not notice any impact to the way your operations are running on flashback will
also document how a fabric interconnected fail. And we'll talk about in the document how both fabric interconnect have copies of each other's configurations so that if one dies the other one continues running your data continues to process. And then when that fabric interconnect is replace the copies of the configuration from the one that died and Voila, you're up and running with both fabricator connects.
The same technology also applies to MDS switches and also applies to our top of rack switches. So you can have multiple failures of a flash back system and still your work continues to process your reports get done. You keep the executives in your business happy. Now let's talk about some of the solutions that we have tested validated on flashed act within
the past year. Let's talk about virtual desktop infrastructure which is a solution that gained a lot of prominence when Covid occurred. Right. Covid hit all of a sudden many of you had to find a way to document and be able to provide a way for all of your employees to continue working perhaps out of their homes.
And so virtual desktop infrastructure came very key as a solution. When we look at flashback, we want to do a number of different projects to test flash stacked across beady eye solutions. So we're looking primarily at Citrix and Beyond the Horizon which are the two most popular VD I brokers in the market today. And as you can see from the chart, we test on a range of hypervisors so that we hopefully can
document the various combinations of VD I broker with hyper Visor that you would like to choose to run on flashback now. One of the latest designs that we've done scales up to 5000 users and users of the latest VM ware Horizon Soccer, which is version eight. You can see in the design here we've got four server chassis of RM five technology with two
of our fabric air connects to MBS switches and two of our top of rack switches. So as I showed in the previous chart we've done all the failed over type of analysis but we've shown scalability to 5000 knowledge workers and knowledge workers is a use case that the type of employee, much like you and I are professional workers who would access a number of different applications to perform normal types of business activity.
This same solution has been documented to 6000 task users think that task users or someone who works on a helpline perhaps Xfinity who you know only access is a couple of different applications to be able to handle calls from the helpline chapel application. So their extensive use of applications isn't as strong as a knowledge worker but still it's a type of use case that many of our clients that are running B.
D. I. R. Sizing systems for. And so we will be able to test with that particular use case as well. Now these are the types of basic designs that we do for flashback around the V. D. I. Space. Now, as you may know, we do have a new design that just came out to market a couple of days ago posted that uses X series technology and so
you can replace the B series chassis is in this design with X series and you'll see a Cisco validated design that test with X series as the key compute platform as part of flashback. Now for oracle customers. Just this last month we posted a new Cisco validated designed with oracle rack 19 C and we tested this running on VM where V sphere seven as the hyper Visor. So here you can see we have two X series
chassis. Um each one is loaded with our tv series blades. Excuse me. And so these blades have been split across two different uh, UCS classic chassis. That way we can implement a complete chassis failure and still show how flex pod is uh, is has the ability to continue to stay up and running even though that complete chassis has failed.
We scale up to eight server nodes this configuration because eight server nodes in an oracle rack configuration we really need to need to about 99% of our customers like yourselves in the marketplace. So through this Cisco validated design you can show you can scale to incredibly large databases using oracle database on VM RVS Fair seven and show how flashback can give you the level performance that your business required
not only today but into the future. Now we realize that many of you out there don't run oracle, you actually prefer Microsoft sequel server and 2019 is the version of sequel server that is most common in the market today. Here we've tested with red hat as the operating system using envy me over Rocky V two.
It's the latest protocol that we've used with flashback and in this particular design we've actually tested the ability to serve a consolidation of multiple sequel server databases consolidated using hyper V as the hyper Visor of choice. And this Flash stock configuration is really ideally suited to those customers who are using for departmental use. Small enterprise-based customers who are
choosing to run Microsoft sequel Server 2019. A lot of customers are looking at 2019 today because older versions of sequel server such as 2008 and 2012 either have gone into Life for support or about to later this year. And so it's forcing a lot of customers to take a look and looking at Harvard platforms as they look to re host and move to sequel Server 2019, flashback continues to be an excellent choice. It is one of the top workloads on flashback
historically and continues to be in County. You're 2022 and we project will be in 2023 as well. Okay, now some of the future plans, well, let's talk about sequel server single server right now in the labs, we are testing sequel server 2019 with flashbacks where we're using the new XL 1 70 array which is really designed for database workloads in mind.
And in this case instead of doing server consolidation, we're going to do data warehousing and we're gonna data warehousing on containers, which is really a new avenue that a lot of customers are going down utilizing containers to be able to house these workloads running on the flash tax system and our plans for this CBD is that we're gonna be talking about some of the testing results in our presentations during Cisco live in the pure
booth at Cisco live june 12th through 16th in las Vegas. And then we'll document them in the CBD, which we will publish in the july time frame. And so you'll see that the configuration here, we're gonna utilizing X series and we're gonna be utilizing the latest components across the flash stock infrastructure so that customers like yourselves can see how the latest and greatest technology,
not just from Microsoft buffet appear in Cisco all combined together to really give you a leading solution to drive your sequel server workloads. Now, one thing that we're looking at, particularly when you look at V. D I is hybrid cloud testing scenarios and so as we move forward, we're using flash tech as the base platform to exercise a number of different hyper cloud
scenarios that we'll be testing and documentation. Cisco validated designs. Now, obviously we're period familiar with running V. D. I on premise on a flash back solution today and we'll be testing how okay you can add another flashback configuration,
expand the footprint within your data center to house virtual desktops. But many customers like yourselves looking at hybrid cloud and so first we'll look at, let's do disaster recovery, let's just say like I live in the state of California, we have a lot of earthquakes unfortunately. And so disaster recovery becomes key to all the customers in the state of California and so beyond the backup,
your infrastructure and all your virtual desktops to say Azure or AWS or google becomes very key. And so that's one of the use cases will be testing but we're also going to be testing cloud bursting in which you need to suddenly add a number of different virtual desktops, the highest covid breaks out again and all of a sudden you need to be able to find a way to host these virtual desktops quickly and easily.
You exhaust your system resources in your data center and so you will automatically be able to set up these new desktops within the cloud. It's called cloud bursting and it gives you a additional level of system resources that you can use when you have those emergency scenarios when you've got a deal to put up desktops quickly and easily. We'll also take a look at data tearing and data tearing is where you're going to be trying to
mix and match warm data and cold data. The best example I can think of is in the hospital and if you go into a hospital you get treated, they keep your medical records within the hospital itself for about six months that's considered hot data and they keep it locally because if you have a follow up visit or another episode you may go back in that same hospital they can immediately access those
records but once you have been in the hospital in six months or a year, they transfer that data up to the cloud. It's called code data. It's still accessible if you go back into the hospital. Typically they're not anticipating you coming back in the hospital anytime soon. But insurance companies need to be able to pull medical records from both.
The warm data that stays within the hospital as well as the cold data up on the cloud. And so being able to do this data tearing pulling from both hot and cold data is a third use case for hybrid cloud that will be testing with our Cisco validated designs in the months ahead. Now, one other thing we're gonna be testing and we're gonna start with D. D. I. Is the exit 4 40 P.
Which is a new compute node for X. Series that houses GPU s. And it can house NVIDIA A 16 A 48 100 GPU s. Or if you have riser being the configuration, it gives you the ability to house in videos. T. Four Gpu GPU s are used in the D. D. I. Space for a lot of applications that deal with highly graphical applications.
I think uh surgical applications. Think about CAM cod auto design, manufacturing architecture. Those types of applications that really demand highly intensive graphic scenarios. So it's a way of offloading the CPU in the configuration from having to manage it themselves. Offloaded to a graphics processing unit. So those Gpus can go ahead and monitor and
manage and create those intricate designs that engineers need to have access to. We call these people power users and it's another use case of D. D. I. They typically demand more system resources. They demand the use of GPU us and our future designs. We're gonna be showing how this new GPU compute node for X series is an ideal
scenario, an ideal solution to the needs of power users in the D. I. Space. So pay attention that in the next six months I think you'll be greatly pleased with the fiscal validated designs for flashback that we're going to bring it to the market and highlight to you as customers. Now. I've talked about, you know, different
scenarios today. Uh Cisco validate designs, V. D I. Oracle and sequel server solutions. Many of you may have additional follow up questions. So you can start by looking at these websites where we'll be able to share with you additional information and you know, please remember as you're looking at who is the best
vendor for you to partner with. You got a lot of different choices. We think Cisco gives you a lot of different answers and answers that many of you have trusted your mission critical applications too because we have Cisco helped to bridge you to where you want to go in the future. Thank you for your time today.
You know FlashStack, purchased to run one application, but have you asked yourself what else could run on it? This session is for you! We will provide an overview of the latest applications that have been tested and documented in Cisco Validated Designs covering VDI, Oracle, SQL Server and more. See the designs and key proof points; and get a view as to the hybrid-cloud use cases that will be proven in our future designs. This will provide food for thought as to how FlashStack powered by Cisco B-Series and the new Cisco X-Series can address your future concerns.
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)