Reservation system for the cloud

CiRBA's Andrew Hillier discusses how its better to think about cloud computing as a hotel rather than an apartment: Cloud applications move in, stay a while and then leave.

Andrew Hillier, CTO of CiRBA, came by to renew our conversation about managing cloud computing environments and a management philosophy that can make best use of cloud resources and features expected for CiRBA 8.0.

The key, Hillier said, is thinking about cloud capacity management from the viewpoint of the cloud environment. See it at as a hotel, not a multi-tenant apartment complex. Looking at it that way helps enterprises make optimal use of their systems, storage and networking resources. This, he pointed out, often means that a smaller set of systems, software, storage and networking can handle the business requirements of the company. Lower costs and simpler infrastructure configurations can save money.

Why a hotel?

Many companies view their data centers as if they were apartment complexes. Systems are acquired to house a single set of workloads for a long period of time. Hillier says that's not how cloud computing resources are really used. The reason that performance and configuration management tools don't make the best use of available cloud resources and costs are higher than really needed is that these tools were designed around the concept that workloads come into the system and stay for an extended period of time.

Hillier suggests that hotels are a better thought model for how today's data centers are being used. Virtual workloads come in, stay a while and then leave. If resources aren't reclaimed and used to support a different virtual workload, data center efficiency and overall data center performance suffer. Costs for systems, storage, and software would all be higher than really necessary.

The game "Tetris" is a better thought model

The game "Tetris" offers a better thought model. Jobs come in and have to be fitted into the available resources. Once they've done their work, they leave. Hillier thinks that it would much better if capacity and performance management tools would have the intelligence to examine what each workload actually needs to operate efficiently. Then that set of requirements can be used to fit everything into the available systems capacity. This might mean moving workloads that used to operate on separate systems to a single physical host or separating workloads and placing them on different hosts. If an optimal configuration can be found, a company's work can be supported on a smaller number of hosts.

Unlike the game, the workloads that are being introduced into the cloud computing environment are not just randomly falling from the sky. IT administrators often know a great deal about what needs to be done and when.  So CiRBA developed a capacity management tool that is very much like a hotel reservations system.

Deep analytics show where things should go

In our past conversations, Hillier pointed out that CiRBA uses "deep analytics" to learn how workloads work, what resources they use and even the impact of plans for future expansion to transform where and how workloads are placed. CiRBA's technology can then automate the process of instructing systems, virtualization and other management software to move workloads into an optimal configuration.

What's new?

CiRBA 8.0 is going to expand the environments that can be managed. Industry standard systems hosing VMware vSphere, Microsoft Hyper-V and Red Hat Enterprise Virtualization can be managed as well as IBM Power systems hosting IBM VM virtual machines.

CiRBA now offers an API for web services that makes it possible for developers to tell CiRBA specific details about workloads and ask the software to reserve necessary space for them to run. CiRBA also offers a Web user interface for reservations as well.

How about OpenStack or other cloud services platforms?

Hillier discussed the growing use of CiRBA in OpenStack environments. OpenStack, Hillier notes, doesn't have a good way to select a system and place virtual machines in an optimal way. It's largely a "first come, first served" environment. CiRBA has developed a way for OpenStack, CloudStack and a few other cloud services platforms to pass jobs to CiRBA's capacity management system so that the environment can be optimized.

Snapshot analysis

Viewing systems as if they were a hotel chain is a very simple model offering some profound insights. Rather than thinking about workloads being static, moving in and then staying in an apartment, it is rather helpful to think about virtualized environments being like a hotel. Travelers come in, stay a while and then move on. The goal of hotel management is hosting the largest number of travelers in the most efficient manner. CiRBA 8.0 expands the number and types of "rooms" that can be managed.

If I consider today's evolving software licensing policies that allow multiple virtual insistence of a given product to execute on a single physical system using one software license, making as many virtual workloads as possible share one physical system could certainly reduce software licensing costs.

As I've pointed out before, putting more "eggs in a single basket," means that that basket needs to be watched very carefully. If performance anomalies start to appear, it would be wise to quickly find out what is happening and move workloads to other "hotel rooms" quickly to prevent outages. CiRBA can help.

As before, I see that CiRBA's ideas are interesting and could be very valuable to companies having highly dynamic data centers or for operators of multi-tenant data centers. If your company has this type of operation, learning more about CiRBA's technology could be very helpful.