In this edition of Solace Says, I interviewed Solace’s CTO Shawn McAllister to get his thoughts on hybrid cloud data movement. Here’s the 6-minute video, with the full transcript below. It’s a good introduction to the topic, and if you want to dive deeper after watching the video I recommend you read the substantial blog post Shawn wrote on the subject, and watch the video that explains the concept visually. You can find both on our hybrid cloud data movement page.
Well, the enterprises that we talk to are in situations where they have, today, they have applications running on mainframes, on A/S400s, Window environments, Linux environments. And what they want to do is they want to build new applications in the cloud, or they want to move some of those applications to the cloud. And they want to do that for reasons of cost reduction, agility, elastic scalability.
And when they do this they sometimes think of either public clouds, such as Amazon, Google, or Azure, or they’re thinking private clouds that run in IaaS such as OpenStack or VMWare. Or they may be thinking of a Platform as a Service such as Pivotal Cloud Foundry and OpenShift. But they’re starting with an existing set of applications. And in the end they’re going to be running applications in one or many of these clouds. And so that’s what creates this Hybrid Cloud environment.
Why this is so hot is because so many enterprises today do have on premise IT assets. And to get there you’ll have to live, at least for a period of time, and maybe forever, in a situation where you have assets deployed in these multiple environments.
So what we mean is you want to get to a situation where you have applications, your existing applications, deployed on premise. And you don’t really want to change them too much. And you’re going to deploy new applications in one or more clouds. It might be public clouds, it might be on premise clouds. And you need to have all these applications communicate and share information because applications don’t exist in a vacuum. The new applications that you deploy in the public cloud, for example, are quite likely going to need to receive information from your crown jewel applications that are deployed on premise. So you need to be able to gather that data from on premise and share it with all of the applications in a secure and governed way, but share that with your applications deployed in these multiple clouds.
You need to be able to connect different worlds of applications together. And you need to do it in a way that spans wide area networks and is very decentralized. So what you want to do is you need to be able to tap into your existing on premise applications whether they are existing JEE environments, or enterprise service buses, or message buses. You don’t want to disturb them, but you need to take out from there the data that you need, typically move it across a WAN and then make it accessible to new cloud native applications that were developed in a more microservice architecture, for example. So you have the connectivity to the old existing applications and you need to support the new microservice type applications over a Wide Area Network with the challenges that that presents, and typically many enterprises will have to meet to support multiple clouds. So that’s what makes it a bit challenging.
Yeah, so Solace has all kinds of capabilities natively built into our messaging technology to provide very rich management capability and visibility for operations ongoing. We integrate with all the DevOps tools like Chef, Puppet, and Ansible, and we have built in security for encryption, for access controls, authentication and authorization. And the key thing is that those capabilities are uniform, whether you’re on premise or in any cloud. And so the tools that you use to be able to manage, monitor, and troubleshoot your data movement is the same no matter where your applications are. And that’s as opposed to doing it yourself or having different solutions for moving data to different clouds, for example.
We have a few very interesting deployments that I can talk about. So one of our customers, for example, is a big betting and gaming organization. And they have a very interesting use case where all of the odds that they produce for people to bet on are all generated on premise, and yet just before a sporting event their users want to be able to receive these odds so they could place bets. So what they do is they deploy into Amazon, Solace messaging that they scale out horizontally depending on the demands of their mobile users as they come on just before a game when they need to scale up from zero to 80,000 connections. And then after the game where they all disconnect so that they can pull down all of the Solace messaging and incur no costs. Yet, the odds are produced onsite, pushed into the cloud and fanned out in an elastic way to all of their users. So that’s one interesting use case.
We have another customer that does end of day risk calculations in the cloud. And so they produce orders of trades during the day on Solace messaging technology and push it up into the cloud elastically, then consumer services to do their end of day computations and send their results back down on premise. And I know you asked for two, but there’s a third one, if you take a look at Wayne Osse’s video he talks about FA Swim in a hybrid deployment there which is quite interesting.