– Alex Bewley, CTO at uptime software (www.uptimesoftware.com), says:
uptime software’s Top Three 2012 Cloud Predictions:
- Cost monitoring becomes a key aspect of public cloud management. Part of optimizing computer resources accurately means understanding what the computer and storage costs are for cloud usage. Cloud users (specifically Infrastructure as a Service – IaaS – users) are tired of “flying blind” when it comes to knowing their cost position at any given moment. This is especially true in highly dynamic cloud computing environments, such as Amazon’s AWS.
- Cost measurement will start to be considered for private clouds. The advent of cost transparency with the public cloud is beginning to move down the IT food chain. Financial managers are beginning to actively question internal costs and seek solutions to manage and report on them. However, accurately measuring private cloud costs is difficult and only currently available via manual methods.
- Maturing of technologies make public and private cloud balancing easy to deploy. Cloud bursting, dynamic scaling and data sharing between public and private clouds are evolving. Underlying technologies to enable bursting are being delivered, but seamless integration with monitoring and management tooling needs to mature.
uptime software’s Top Three 2012 Virtualization Predictions:
- The Maturing of virtualization management consoles. Managing a virtualized environment has typically been done with several point tools (storage, guest lifecycle management, performance and capacity management, service book). Moving into 2012, we will begin to see a greater integration of these tools, reducing the complexity imposed upon an administrator.
- Improvement in automation and deployment. As tooling becomes more integrated, it will become easier to deploy complete applications and have them monitored and understood by tools. Service Level Agreements will accurately reflect the dynamic state of the application in these scenarios. For example, as components of an application are dynamically re-sized (either up or down), monitoring toolsets need to understand scaling down does not equate to downtime.
- Service catalogue and application catalogue – not there yet. The ability for end-users to select an application to be dynamically deployed in a private cloud is a great idea, but realistically it requires too much configuration and management on the back-end to be viable. Rock solid automation, deployment, chargeback and integrated management consoles are all prerequisites for this, but these integral aspects are still a ways off from being a reality.
Solving some of these immediate challenges, up.time
Follow uptime software on Twitter: http://twitter.com/#!/uptimesoftware