Windows Azure the right way to move your application to a public cloud

Rightscale®inc Solutions Architect. Cloud management enables organizations to easily deploy and manage business-critical applications across public, private, and mixed clouds. Rightscale provides http://www.aliyun.com/zixun/aggregation/14083.html "> Cloud infrastructure and Applications for effective configuration, monitoring, automation, and governance."

When migrating to the cloud, it is normal to generate anxiety. But it shouldn't be. The-it industry has experimented with the basic elements needed for successful cloud deployments: Redundancy, scalability, and automation. These ideas are not necessarily new, but now they can be collectively applied to new Windows Azure virtual machines (as a service infrastructure), which takes less time to achieve geographically distributed, highly redundant architectures than traditional data center approaches. This is an appealing proposition for companies of any size, since 2006, in Rightscale we have helped large and small companies deploy applications in the cloud. Our pioneering cloud classification management provides a platform to enable organizations to deploy and manage applications in a cloud environment, and our customers have launched millions of servers in the cloud.

Rightscale is proud to be a strategic partner in Windows Azure, and as a way to introduce ourselves to the Windows Azure community, we want to share some of the best practices from our unique experiences to make the world's largest cloud deployments more powerful.

Redundancy is kingly

For a long time, load-balanced Web application tiers and mirrored databases in system deployments are a common way to deal with service failures-the advantage of cloud computing is that geographically dispersed deployments enable servers closer to the client to provide resources to clients while providing disaster-recovery capabilities beyond the data center boundary. When you design system redundancy, you need to consider the following points:

• Don't just schedule server redundancy, it's useless if you can't get the server uptime-remember your public DNS infrastructure is also important.

• Plan for service failure and use it frequently. Disaster recovery and failover testing are key to ensuring long-term operation.

• Think about the meaning of automatic failover. For some systems, automatic failover can result in data corruption-testing and evaluating your failover plan to not only run the server properly, but to keep the service running and data intact.

Scalability Opens new doors

Be able to optimize your long-term total cost up or down according to your needs. When your application requires more resources, you can access hundreds of servers from the cloud through several servers in hours or even minutes. When this demand is reduced, you can quickly reduce the number of resources so that you can save the cost of idle resources. When you are designing a scale:

• Remember to separate the deployment to resolve the errors encountered.

• Evaluate the cost of data transfer between regions.

• When you need to increase or decrease the resources, consider the performance metrics and the system triggers will show things.

Automation is Everything

All simple tasks need to be done by yourself--time will be wasted. Building an automated process can help your technical team focus less on production support projects and focus on updates, production, functionality, and solutions that will drive your company's development. If you want to use fast and on-demand scaling up/down, automation is the key to ensuring that you can start a server and minimize the time it takes from booting to a normal working state. For a Web server, this can contain everything from the installation of the admin site code to the connection load balancing settings. You can easily zoom in and out on a mirrored database server, but automating the process of initializing a mirroring session and producing a new level two node will make it easier to expand the data tier. Some things to consider:

• Automated processes for resource allocation and release-both sets of tasks focus on detail, requiring 100% to ensure application accuracy and uptime.

• Learn PowerShell and learn about cmdlets products you can use. Learn about Msdeploy and MSBuild. Use existing tools to reduce the deployment of labor.

• Use this time to improve your process. Find out what best practices you are using the product and try to implement it in your new automation process.

It is not new to take these methods alone. But by doing it right, you can get your team to focus on physical hardware instead of focusing on high-availability solutions, even though your system (or your team) is small. Both large and small deployments benefit from redundancy, scalability, and automation. The more perfect your process, the better your team will grow in the short and long term. If you want to try these basic technologies on Windows Azure, Rightscale

The

Free edition is the fastest and easiest way to start.