Many complexities to ponder before going hybrid
Hybrid clouds offer incredible potential…in theory. But they also imply quite a few new challenges and complexities for organizations interested in using them. Getting best value from a hybrid model will mean thinking these matters through in advance — and implementing an effective strategy to handle them.
For instance, there’s the daunting question of security/compliance. It’s hard enough locking down data and applications within your own architecture, which you own, oversee and manage directly. When data is flowing back and forth between that architecture and a public cloud, things get harder yet. The same issue applies to compliance — complying with government regulations that mandate tight controls over data access rights will clearly get more awkward in a cloud context.
Performance monitoring is also an issue. Most organizations have some form of IT asset monitoring in place, to ensure that key assets continue to work as expected, emerging problems are detected, and (if possible) future problems are anticipated and precluded from happening at all. But that functionality begins and ends inside company walls; there’s no easy way to accomplish the same goal with somebody else’s public cloud.
Workload management/governance also comes to mind as a hybrid cloud challenge. Ideally, you’d be able to oversee how workloads are executing and, as changes occur — perhaps a spike in demand affects certain workloads — you’d be able to respond right away and in a business-prioritized fashion. But here, too, lack of direct insight and control stands in the way.
And then there’s arguably the most basic problem of all: how you’re going to link your in-house applications and data with the cloud.
Perhaps the most obvious approach would be to write new code. But if you work in IT, you already know how many problems that idea implies. New code takes time and money to create, debug and roll out. And there’s always the possibility that it simply won’t perform as expected. If the glue holding your hybrid cloud model together dries up and disintegrates, your cloud strategy is obviously not going to work as well as you’d hoped.
Use what you already have to better business effect
So let’s see if we can come up with something better — a way to solve these and related issues, and do so for the lowest costs and in the shortest time.
To begin with, obviously it would be great if you could leverage the investment you’ve already made in IT management solutions. For instance: monitoring. If you already have IT monitoring tools, and use them every day to good effect inside the company, it’d be nice if you could extend their reach — somehow leverage them for the same kind of purpose in a public cloud environment, too. This way, not only could you drive up uptime and performance, but you could also do it via an interface and feature set you’ve already mastered — all but flattening the learning curve.
Furthermore, that kind of insight could be passed along to other tools that do other tasks. Workload balancing, for instance. Once you can visualize and quantify how cloud assets and resources are being used, you can use that information to prioritize workloads based on business goals.
This way, instead of needing new solutions, and spending more money to get them, and time and energy to learn to use them…you’re really just using the solutions you’ve already got in a new way.
Already we’re making considerable progress toward rendering the hybrid cloud model a lot more optimized for business purposes.
In the area of security, of course, you do still face the problem that the cloud owner has direct control over the cloud’s capabilities — they choose the security solutions, they do the configuration and they handle the ongoing management. But what you can control is the data. Specifically, you can synchronize data directories so that there is a perfect reflection between the on-site architecture and the cloud architecture. For quite a few everyday applications, like e-mail, this is a great way to take full advantage of cost-effective, pay-as-you-go cloud pricing without creating a lot of new security hassles and complications.
Last, let’s tackle the worrisome issue of data and application integration between your in-house architecture and the cloud. Imagine that, instead of expensive and risky new coding, you could take advantage of predefined templates and technologies that create the necessary link between the two environments — maybe tweaking it here and there, as necessary.
For instance, you might find that to use Salesforce.com (a cloud service) you need to route information from an SAP application running inside your own infrastructure. To do so, you would just use a template. It would contain the necessary logic to perform that routing, based on its awareness of both SAP and Salesforce specs.
Now, obviously for different applications or clouds — a lot of possible combinations — you would need a lot of templates. But it could be done, in theory.
IBM’s hybrid solution drives down costs and complexity for better cloud performance — easily
IBM has made that theory a practical reality.
Today we offer a hybrid cloud solution, based on Cast Iron technology, which actually is comprised of hundreds of such templates. You can quickly and easily use these to orchestrate information flow to and from the cloud with low (or no) custom coding necessary.
The story gets better yet. Suppose you have a significant investment in IBM Tivoli solutions, like IBM Tivoli Monitoring, IBM Tivoli Netcool Impact or IBM Tivoli Service Automation Manager, and you use them to manage internal services. Now you can use them to manage your cloud services, too — broadening their reach and multiplying their business value.
All you’ll need to do this are some service management extensions we’ve designed specifically for hybrid clouds…and these can be downloaded from IBM’s site for free. What’s more, the next generation cloud service automation solution – SmartCloud Orchestrator built on OpenStack technology also uses these easily available, pre-packaged extensions to support hybrid cloud and take your cloud services to the next level.
How’s that for a smarter hybrid cloud?
Discover more Smarter Computing Breakthroughs posts:
- Cloud Storage
- Stream Processing
- System Interconnect Improvements
- Middleware Optimized Systems
- Information Integration, Pt. 1
- Information Integration, Pt. 2
- Unified Management
- Data Security
- Image Management
- Cross-Platform Virtualization
- SMP Interconnect Fabric — Simpler, Faster, Smarter Scalability
- Intelligent Threads: Tuning the POWER 7 Processor to your Workloads
Marco Sebastiani is a Product Manager at IBM Cloud and Smarter Infrastructure, being responsible for the definition of the product development, go-to-market and investment strategies of the Tivoli Cloud Portfolio. In his role, Marco works with worldwide enterprise customers to help them use cloud technologies to solve their business needs. Marco has 17 years of rich experience with IBM, covering senior management positions in product management, product development, investment strategy, working with worldwide enterprise customers.
To effectively compete in today’s changing world, it is essential that companies leverage innovative technology to differentiate from competitors. Learn how you can do that and more in the Smarter Computing Analyst Paper from Hurwitz and Associates.