Datrium VMware Integration
Transcript
[Music] hey how's it going with Keith Bowser from the CTO of Iser calm and you're watching the CTO sponsored by Bayes rule we're on premises where we've been doing a series of interviews if you follow this at all we've been at VMware talked to their product staff and good from the cloud bu all the way to the networking bu and we're also also talking to the ecosystem of partners from VMware which one of them is de trium who's sponsoring this video so Andre thanks a lot
for joining us and Mike I don't know either you on the CTO advisor before so Mike go ahead introduce yourself so thanks Keith Mike McLaughlin I'm the director of technical marketing here and I basically head up a lot of the solutions that we test and validate run our demo labs do a lot of our reference architectures white papers I get to play with the product a lot okay I'm a VP for solutions and alliances so I'll discharge you around partnerships what we do with those partners
how many partners we need how deep we go with them the work we do with VMware s ap Oracle and also solutions and we part of the same team so let's talk about specifically one of those lines I counted my curiosity I Pete when you said SAPT gentlemen I'm a old school SAT guy my last full-time job was a safety infrastructure architect but with that said ecosystem is critically important when you're talking about HCI open converged there's vendors out there trying to create their own ecosystem
and then there's you guys who are more of your like data protection storage cloud is key pillars of your solution and then you integrate with partners let's talk about your integration with VMware where do you innovate and add value in VMware say yeah no that's a good question first of all it's important to understand that customers only buy infrastructure because of the applications right no there's no applications there's no infrastructure right so we need to leave by the applications so we need to make sure that
we work with them we perform for them ultimately they serve the users so we do support VMware as a hypervisor with that we have the entire VMware ecosystem we not only support but we integrate into VMware you have that the single experience from a operational perspective so when you start talking about effects and no mitigating the number of different people to manage the infrastructure we solve that piece the way we innovate with VMware is when able VMware to scale where they have never been before so
some of our performance benchmarks that you will see out there scaling storage in VMware 228 servers and performing in a way that nobody has done before and the way we do that is no technically speaking up with the ability to scale compute independent of 128 notes you're the marketing guy you're the technical marketing guy but how do you prove that this that's rather large like a visa no visa has a great solution going goes up to about 30 nose I think right 128 sounds like it's
a lot it is that it's a pretty big cluster it's a pretty pretty big infrastructure right to run applications on and we actually did that one live with some other partners we went and built one I mean our systems growth sits 128 not a theoretical we've racked up a bunch of servers Network them started them all up ran different workloads and really pushed the boundaries of not only the compute level but also the storage level so you can scale out compute 128 nodes to get the
applications running we also can scale out our back-end to 10 data nodes so we actually put that all together as well it's a fairly large investment in in our time resources and a couple of our part is out in a another large state in the Union where we set that up and tested it the results are posted but yeah we had to build one of those I mean that's kind of what we do on this side is prove this out we'll build even small-scale one host
all the way up to large scale full-on DB X's so let's talk about practical use cases for a lot of not just scale but a management release under I love the fact that you said that infrastructure only exists because of the applications so let's talk about application specific instances what are customers finding the most use for dasyam dvx or what are some of the pain points in specific applications that date from dvx solves Oh personally I'd love to hear the SI p story because I can
easily relate to that yeah the way you think about applications and business outcomes is you need to remove the pain to focus on on the business right so we need to remove infrastructure pain and know many vendors at all we automate these we do that the real truth is that the Kirk's is performance the the biggest cause of applications failing over time is performance so if you remove that that's the first step right right so we know that the hm performs well and no more than
any system out there and can scale to know this amazing performance numbers that I don't even know any application that would require all that to run today the second piece is okay we know that we can do that now we need to go invalidate all this application you see many of disk you know business critical applications or live critical applications they try to run in converged infrastructures before and they failed because performance was not there and I have ability was not there and there was no
features like snapshotting in there certain aspects that need to be there and any look thinner talk to us a paean talk about Hana or you talk to you Epic Systems and there are others they don't support convergence today because it failed in the past right so now we are in that momentum oK we've got a great solution but some of those core applications are still not supported in conversion conversion architectures are all the same and I don't blame them you know the it failed in past
so now we are in that process okay we have something new we have something more innovative we have something that address performance address is scalability address now is not shorting data replication data management we need to redo it now that said I'm talking about a few applications out there that require all that for the large majority of applications it's fully supported today in Converse and convergence architectures know as as for sa peewee you know they support one business suite on VMware with daydream now epic will
see epic supporting hyper space and VDI on on date room and all the platforms right well and there are other other applications out there now the use cases from a verticals perspective it goes like from now finance oil and gas in healthcare to media we now we have for healthcare have PAC systems there's a large data capacity we have media companies using for video streaming so it's the broad range of use cases that we cover today so we've talked a lot in the past with you
go in to Chara about cloud DDX and in health integration goes along that let's talk about the eCos implication zoom how does something like taking datum dvx extend out to a cloud dvx technically when you're talking about application support where's the value of a cloud dvx in that ecosystem yeah so the cloud zvx is is is our system that has been visualized and put it on a de Bresse to start with the second piece is we have what's called we call global deduplication so data is
never duplicated in the painting where they where the dvx is running if it's on pram or in the cloud so now you can imagine that you make cloud backup and cloud BR and no and be able to birth data to cloud so you can really in like in the dr scenario you need to have that deduplication across your environments otherwise you moving terabytes of data and that's just not gonna work right so you want to make sure they move only one bit of data if required
what cloud edx does today is we enable you to replicate all your data with snapshots to AWS store they're in a very cheap way manner and be able to restore that whenever you need in your site or in a dr site that's the first stage you can foresee as using that data to instantiate also be are in the cloud okay so before recording we talked a bit about counting the work you guys we're doing with vmware from a paper or research perspective can you tell us
a little bit about the work that you guys did combined with vmware yeah actually Mike was responsible for hundreds done this before too so you know the recently some one of the applications that we looked at was virtual desktops horizon there's actually two that we've seen in our customer base that actually sort of take advantage of the platform when you look at the applications databases and desktops databases have high latency dependencies and putting flash on the host makes those just scream desktops have the user experience
right and the more users you pile into a virtualized environment the less you want to worry about the storage i/o all right so our architecture puts that up on the host so flash on the host kind of takes storage i/o off the table so we were able to look at building out how you would scale that application in an environment from just a single host yeah potentially up to 128 but let's just say up to something like you know a couple of thousand users all right
and look at not only is just the the vetting of the testing and making that work and hitting all of the goals of user experience and you know time to do certain operations but also how the platform grows as an as a customer might so as you add more users you can add more hosts okay the flexibility of our platform actually lets you sort of pick those hosts to suit the needs so there's a lot of simple things in the way the dvx is set up
that we pushed out in this reference architecture for horizon so what we wanted to do was give our basically give our customers kind of the line in the sand and says if this is what you're looking at here is what you can expect to get and then here's where you can go if you need more horsepower or you need like GPUs or you need more density here's how you make those choices so the reference architecture it's just that it gives you a reference point of a
particular application on the on the dbx plan well here's what it's interesting this is not a just datum paper that really this is a joined datum in VMware work which proves that we can't coexist we can't work together VMware has no V saying we've got our own product right you know and we can coexist the the ecosystem that vmware vsphere drives is just enormous is gigantic and is awesome to be part of that yeah yeah I love the story the a lot of times we'll get
into this debate about markets that VMware gets into and whether or not that competes with customers but I think van where show repeatedly that you know what you can take our solution but if you want to build on top of V spear and then move out from V spear we're perfectly okay with that I think that the combination of date um dvx V spear and horizon shows that development to the lines if I really appreciate the work the mic the you you and your team is
done with VMware I mean II just to dig in pool of bits and pieces of Martin you know pick at it a little bit I like like challenging you guys but good you surprised me with 128 no yeah piece up I'm glad I wasn't there to pointing fingers I'm glad I didn't have to haul all the equipment out and back it was truckloads for a couple of days well that's it for this CTO dose if you want to find out more about datum you can visit
them on the web ehm comm Twitter handle for a startup is pretty simple at Twitter me and atrium on Twitter you can follow more about the CTL visors seq adviser Tom maurices said Kemp subscribe to the blog podcast and of course the CTO daily dose talk to you next dose
