AWS EC2 inside your data center
Transcript
[Music] hey how's it going his keys Hauser from the CTO viator.com wolf another CTO dose it's been a little while started the job at VMware I've just been busy a lot of you hoped that I continue these CTO those conversations and I very much will and it's not gonna be all about VMware because that's not why you came to see me we're gonna talk about the significance of AWS is announcement that they're gonna run ec2 instances so snowball ish so this video might be a little
longer than the ones in the past we'll walk through kind of the history of the snowball itch and talk to why to get to why this is a really significant development in AWS is product portfolio and why it makes sense for the industry and where the industry is going so obviously for the VMware cloud on AWS perspective that kind of makes sense now but not just because I'm being more employee being able to run VMware workloads next to your AWS workloads reality of where we're at
and hybrid IT there's a need for it the opposite is the case as well the ability to run ec2 workloads or ATV c2 and AWS services and your data center is also a need we're slowly getting there so let's walk back the history of the snowball edge years ago AWS introduced the snowball pretty straightforward concept if you had more than a couple of terabytes of data that you needed to get to AWS the best way to get that fare is via physical hard drives snowball allows
you to rent a hard drive or a set of hard drives from AWS copy the data high speed into your data sent from your data center into the onto the snowballs and get the snowballs over to AWS you can't beat a truckload of bandwidth don't underestimate a truckload of tapes to transfer large sets of data very very straightforward you know you can even extend that idea all the way to AWS is snowmobile which was announced a couple of years ago side by side with the snowball
edge which if you have hundreds of petabytes of data then you get a basically a semi truck full of hard drives to transfer that data or simplifying it but that's the history of how do you transfer data into AWS so AWS came out with this concept of the snowball edge so what was the edge is snowball these are basically servers for to you or more so servers to 1u servers with hard disk in them which primarily allows you to say okay if I have a let's
say a continuous need to transfer data from eight from my edge the classic web Mille the poster child use case and this is not a one-time data transfer how do I continuously do that well you take this HCI looking solution I call it an HCI looking solution because it's a four note solution with a expanded file system that allows you to copy data to that cluster of servers once you've copied the data you can take one of the units out ship that to AWS import the
data while you're continuously copying data so this gives kind of a landing expand opportunity well what if you need to run compute against that data let's say it is again the wind farm and you want to process IOT data and you want to be able to say you know what if a sister goes off and it's outside of reading I want to kick of a process well you know what that sounds a lot like green grants AWS green grass in AWS IOT solution well green grass
and which is AWS is IOT solution and lambda which is ata ws is functions as a service solution perfect combination kick it off now the nelsons that happened a week ago or so ago what if that process was a little bit more complex than saying okay we need a lambda function that's going to kick off a chain of events on some higher-end services whether those services are in your private data center or in AWS cloud or some control systems inside of your manufacturing floor or your
wind farm what have you actually needed some fairly robust compute for Linux or Windows instance to do that control I have a for you HCI looking system on-prem already right why can't I just run a full am I on that solution well that's where the new announcement comes into play you can now get super small ec2 instances on those edge devices which enables you to have this distributed IT or a distributed cloud where AWS still has the AWS control panel in the AWS state data centers
but you have ec2 capability in your edge wind farm or data center so is this a threat to you know Microsoft Azure VMware's cloud on AWS VMware vSphere solution overall not today you're still missing an awful lot of cloud capability that you need there's no database services at the edge you can't have cloud without having database services you're these are really really small instances so if you need serious compute if you're doing you know if you want to do some high-end machine learning training algorithm training
at the edge these systems are not really equipped to do to do that this is basic functionality however this is where the industry is moving we are going to need cloud capability inside of our data centers when I say cloud capabilities I mean public cloud Microsoft Azure stack makes sense if you're a Hall microsoft.net shop and you've embraced Azure everything the same thing for AWS ec2 if you've embraced AWS and you want AWS everywhere this new ish compute solution makes sense power very few of us
are all anything we're not all Azure we're not all AWS we're complex environments that need a combination of these solutions that drive business value I hope you've enjoyed this you want to learn more about the CTO advisor you can follow me on the web the CTO visor com1 twitter at CTO visor and of course you can find the CTO dose videos CTO dose comm talk to you next CTO dose
