January 11, 2016 | Written by: Dan Kehn
Categorized: Events | Hybrid Deployments
Share this post:
HybridBlueBank, a hypothetical business, provides a public cloud-based mobile application to help customers manage business loyalty points and provide feedback. HybridBlueBank service reps receive customer feedback through an application hosted in a private cloud with the intelligence to analyze text for expressed sentiment, giving a clear prompt for a response.
In this webinar, IBM Bluemix Technical Product Manager Jeff Brent discusses Bluemix Hybrid, the cloud platform as a service upon which an innovative business like HybridBlueBank depends. Jeff demonstrates the Relay technology that keeps Bluemix software up to date and in synch across public and private cloud deployment types.
Watch the Webinar
Questions & Answers
- What would be the first and quickest steps to take in transitioning from traditional enterprise development to cloud native? Get innovative juices flowing across your organization because you’d be surprised what the types of solutions or ideas that are in your own building. And we’ve seen a lot of our customers that have started with a hackathon, to really get going with the ideas, and push that innovation envelope a little bit further.And you can use that very quickly and easily in a public environment. And you’re paying pennies on a dollar for these resources. You’re not going in there and committing long-term contracts or buying perpetually licensed software for years and years of use.
You’re able to quickly get on the platform, create applications, deploy them, and get started in a very rapid fashion. And then when you have a lot of ideas, what I think you need to do is, take a step back and understand what types of workload you have, and compare those to your security and compliance requirements and standards.
With all three deployment models, we have an opportunity for you to address any type of workload, whether it’s a multi-tenant public cloud, or a single-tenant private cloud that’s still hosted offsite, to also having an opportunity to host these cloud solutions inside your own building.
So what we saw was effective with even our own adoption with Bluemix, is to sit down and figure out data is going to be allowed within these applications in these various different zones. And that makes it very easy for when someone has a very innovative idea, to come forward, you have a system by which it’s consistently putting that workload in the right place.
So starting off in the public cloud and doing a hackathon are some of the ways that you can quickly get started. And then find the right entry into your cloud journey. We provide the options for you to do it in any way that you see fit.
- How does Relay work if I have a dedicated instead of a local Bluemix deployment?
It works exactly the same. It’s just that, we have a Relay for our local environment. It’s something that we branded the connection between the CICD pipeline and your own data center. Now in the context of dedicated, the Bluemix is still in the SoftLayer, but it’s still an isolated, private account.So the Relay works just the same. It’s just that we have the ability to hook that to the CICD pipeline. All of the environments, whether it’s public, dedicated, or local, run off the same kind of assembly line. And then we use that same Relay technology to go from our source code repositories, into our staging environments, and then finally through coordination, and some of our other internal tools like rational team concert, are then rolled out to the various different instances of Bluemix, around the world.
- If I’m a developer, is this something I need to be worried about? Or is that only something that administrators are ever look at?As a developer, you never need to worry about Relay. That’s the beauty of it. As a developer, you’re really focused on your applications and not necessarily the infrastructure. And even the administrator persona doesn’t interact directly with Relay in any way.But what they do interact with is the tools that it’s enabling – the catalog management, interacting with the security and maintenance scheduling aspects – because that’s how the platform is kept up to date and consistent. So you always know you’re running with the best platform that’s available, and that your applications are deployed with the right security fixes and everything else in place.
That’s really from an administrator perspective. You don’t even interact directly with Relay in that role. What your doing is you’re interacting with the benefits of it.
- What happens with a problem in the platform, for example, a service disruption? What do we do in terms of the user experience from my application?If there’s a problem with the service, then we would receive an alert. So there’s operations’ tasks that are involved there. And when we receive an alert from our operations side, just like from the public perspective, we would intervene and make sure that it’s brought up to date.So notifications would go out to the users of the platform. And we are even working on things like web hooks to tie these types of notifications or events into existing incident management systems.
- Does that come into play, though, if you have multiple locations for your Bluemix? Would that also help it be done automatically?Yes. So actually, we have the ability to run your applications. Some or our customers are using multiple instances of Bluemix, and then leveraging a global load balancer to distribute load across it for continuous availability purposes.
- The next question turns the table to the developer’s viewpoint. It looks like a package solution; however, it’s similar to Agile. How so you distinguish this from Agile?I wouldn’t distinguish it from Agile. It is quite agile because that’s how we run our business. So what we’re enabling is really the adoption of Agile through providing the platform, not only for you to innovative on iterations and sprints, but also our continuous integration, continuous delivery pipeline from Relay is keeping the platform and providing features from us, at a very regular basis that’s based on the Agile methodology.The Bluemix Garage Method was introduced with much fanfare a few weeks ago also speaks to this point. Agile is part of that. You’ll find the Bluemix Garage website talks to Agile and how it fits in with the rest of the product.
- We have a lot of IBM systems at our data center that we want to repurpose when we move to a private cloud architecture. Would Bluemix allow us to do that? And what would be the extra effort that’s involved?So I can interpret that question a couple of different ways. So I’ll answer it both ways. If you have a lot of systems, then you have a lot of intellectual capital that’s built up, and that’s great. What we want to do is provide Bluemix Local or these solutions, even dedicated, to interact with your existing assets, to interact with, like I said, your incident management system.Take advantage of the APIs that you’ve created through your SOA initiatives to interact with your existing tools chain. So in that case, Bluemix Local kind of fits in to be this platform that allows you to leverage all the things that you have around it currently, very seamlessly, and you can just get on that path to innovation much faster.
The other way I can interpret the question is that, I have a lot of systems, and maybe I’m thinking about repurposing it from a hardware perspective. And, in that case, what we do is we provide, like I said – for local, we provide a systems specification for VMware.
So if you have infrastructure that currently exists, then you can repurpose that infrastructure if it’s based on VMware, and make it available for a Bluemix Local installation.
- So if you have something that’s hard to move or you don’t want it to move, for some reason, you might choose to go local. But, otherwise, you would leverage a dedicated or maybe even a public cloud. Is that accurate?Yes, that’s accurate. What we see a lot is that if our customers have an existing service-oriented architecture strategy, you can get and work on that very quickly. In other cases, they may not have that strategy in place, so then we can offer an API management like solution to start tapping into that existing infrastructure or solution set, exposing its functions, minimizing its investment and risk, and eventually scaling down its footprint. And you can provide more and more engaging experience on the Bluemix platform through our applications.For example, if you have relational database with APIs, you can innovate in the future, while still retaining legacy systems, if necessary. We have API management where you can use an API management with policies. There’s also the capabilities to add new service brokers to this environment, so you can take advantage of and maybe implement your own service broker to bring those assets into the catalog.
The key is getting it into your developer experience so that it’s consistent, and those methods help bring that into and provide that function right next to either the IBM created services like Watson and others.
- What is the uptime guarantee that you offer – something like 99.99 percent or other?We just recently released the SLAs for public, dedicated, and local, and the SLA is 99.95, I believe, for multiple instances of Bluemix. So, like I said before, we have our customers that may want to leverage or deploy an application to both Dallas and London, in a public sense.We provide a higher level of SLA, when you have your applications deployed to two regions. And the same counts for our dedicated and local environments. And you can use a global load balancer. Today, we’re actually working on some stuff to make that more easily consumable from the platform. But with deploying your instances to multiple locations gives you that availability that you’re looking, the most continuously available.
- What happens to Bluemix Local if the data center where it’s hosted, experiences a network outage, and loses connectivity to Relay? What if this network outage lasts for hours or even days?We certainly detect that an outage occurred. And then from our operations perspective, what we do reflects our a very collaborative relationship. We would contact the customer, and find out what’s going on, and see if there’s anyway that we can help, in order to get the systems restored because we’re working together on it. So if Relay connectivity is dropped, then something must be wrong, and we would work with the customer to fix it.It has no impact as to whether or not your systems or your applications will run on the premises. But it would hamper our ability to help support the platform and make sure that everything’s working properly. So we would reach out and get that problem resolved.
- Let’s say we are very lucky and our Bluemix based app gets a lot of user attention quickly. Wouldn’t that be nice! How do I scale up my ability to serve my users on the fly?We have an auto scale feature as part of our Bluemix catalog. You can automatically scale up your application, based on rule sets that you apply through our scale service. In the case of public, I mean it’s practically limitless. You might have to take a look at deploying your application to multiple geographic regions.In the case of Dedicated and Local, again you have the elasticity of a SoftLayer data center. So the popularity of an application would only increase your ability to take advantage of that elasticity. In Local, it becomes a little bit more difficult, and that’s the advantage of having the dedicated and public models.
But in Local, the customer is providing the infrastructure or the service. So if your popularity exceeds your ability to support the application, then under the circumstances of Local, you would have to get to the point where you’re racking and stacking or reallocating the virtualized IS, like VMware or OpenStack.
- I get that cloud computing can bring services to two guys in a garage, but how does a small company manage that part of their IT budget for cloud services, for their launch, and going forward as their business grows?We have a calculator out there for you to kind of do estimates and we’re interested in the feedback. If the questioner wants to go and take a look at, under the pricing tab and public Bluemix, you can go and use the pricing calculator to try to determine.We’ve been working at several iterations of that calculator experience to try to make that point – if you’re going to be using this, what would that cost you? If you’re going to be using that, how much would that cost you, and guide you through that experience. So we’d be interesting in hearing your experience if they were to try that public calculator.
There are also features in the platform where you can set spending alerts. Once the application is up and running and your using services like our scale, then there’s a feature and function where you can set alerts to let you know when your system or your application is using a certain amount. I’m not sure if it would be considered a budgetary tool, if you will.
You don’t have to pre-provision this capacity in public. In the dedicated and local models, it is based on capacity. But you can also expand and contract that capacity, based on your usage. And the administrative view that I showed you, with being able to see how much of the application data pool was left or DEA pool was left, it would help aid in that.