Formed in 2009, the Archive Team (not to be confused with the archive.org Archive-It Team) is a rogue archivist collective dedicated to saving copies of rapidly dying or deleted websites for the sake of history and digital heritage. The group is 100% composed of volunteers and interested parties, and has expanded into a large amount of related projects for saving online and digital history.
Every day, more and more organizations take the plunge and migrate to the Amazon Web Services public cloud. What they leave behind is the cumbersomeness of the traditional data center, but despite the cloud’s benefits in scalability, agility and efficiency, they discover a new set of challenges that need to be overcome.
Gartner estimates that more than 50 percent of enterprises will have adopted a hybrid cloud approach by 2017. Transitioning from a traditional on-premise IT infrastructures to a public cloud can be daunting, however, and success requires a different mindset and range of skills. This shift can be particularly challenging for traditional enterprises because they have a lot at stake. Here are the five main challenges in moving to the AWS cloud.
Cloud Finances
Different organizations have different financial approaches, and their choices of IT infrastructure reflect this fact. For some, the on-premise approach of making a large, upfront capital expenditure to purchase infrastructure and then capitalizing the investment over time may be the favorable option because they prefer to keep complete control over their IT environment. However, for others, a heavy initial expense is not ideal, so a cloud approach with only ongoing, operational costs is more fitting. This option may be particularly suitable for organizations with fluctuating needs on a monthly basis, as an on-premises data center will not offer them the flexibility they require.
Regardless of the approach, it is important to compare the respective costs before deciding which one is most suitable. The best option may be to combine both on-premise and cloud to create a hybrid cloud environment. This will allow for steady workloads to be kept onsite while bursts in demand can be processed by an on-demand, public cloud.
Security, Availability
The idea of handing over all of your data to a public cloud provider can be a daunting because of the obvious security and availability concerns. However, public cloud providers must adhere to strict compliance protocols and can implement and maintain much higher security levels than on-premise installations because they have more available resources.
In fact, enterprises have avoided migration to the cloud for a long time due to security concerns. However, this is slowly changing, as IT leaders now realize that running on the public cloud actually can bring the advantage of greater security because AWS has industry-compliant cloud infrastructure (HIPAA and PCI-DSS standards, for example).
Nonetheless, AWS customers must understand their responsibility in protecting user data and implementing a service level agreement (SLA). For more information, take a look at the AWS “sharing the responsibility” model.
Rehost (Lift and Shift) vs. Rebuild (Re-architect)
Migration to the cloud requires some amount of preparation before in-house applications can be moved. Gartner describes five different methods of application migration, and two of them are specifically relevant when discussing IaaS—rehosting and rebuilding.
Rehosting (also known as the “lift-and-shift” approach) involves taking an application and re-deploying it to a different hardware environment without making changes to the application’s architecture. Although this is a quick-fix solution that will have your application up and running on the cloud quickly, rehosted applications probably will not be able to utilize the main advantages of the cloud (such as horizontal scalability) because they were built on top of custom vertically integrated hardware. Rehosting allows organizations to get away from high on-premises infrastructure costs, but the approach is suited mainly to applications with easily defined patterns.
Resource-intensive apps (such as those used for image rendering or big data analysis) are more suited to the rebuild (or “re-architect”) approach, in which the original code is scrapped and replaced with new code. These types of applications naturally rely on commodity hardware and have the ability to scale, so keeping them in closed, on-premise silos could mean you end up with performance issues and high costs.
Rebuilding is naturally a time-consuming and costly process. The solution, therefore, lies somewhere between the two approaches, with the main deciding factors being cost and security of the workloads.
Get it Right from the Start
At the beginning of the process, be careful in selecting which apps or services you move to the cloud. If the first few are successful, that will set the tone for the whole migration. If some of the early projects fail, enthusiasm for the migration will dwindle and make it more difficult to complete.
Be smart by choosing projects that are most likely to be successful, such as non-critical applications or services that can be migrated easily. Make sure your environment is transparent from the start by using modern monitoring and logging systems that can collect large amounts of log data and then parse it to find events and provide insightful information.
In addition, knowing your cloud environment will significantly increase your success rate. Don’t get too ambitious and try to do it all at once. Follow each small success with another, and your team will ensure a smooth migration.
Beware of Staff Apprehension
Any dramatic change within an IT environment will face staff apprehension. You might expect the shift to be met with resistance from corporate management, but the real doubts will come from your IT team. After working with on-premise infrastructure for years, administrators will not welcome any changes they think could jeopardize their jobs. Help your team embrace this shift by making sure they receive suitable training to prepare them for a new set of challenges.
Log Analysis and Metric Collection
Organizations moving to AWS suddenly find themselves managing a highly scalable and highly dynamic environment that requires a new type of log analytics and metric collection. The need for centralization of data is crucial in dynamic environments because often you find yourself trying to troubleshoot a problem on servers that no longer exist.
During the most recent re:Invent, AWS CTO Werner Vogels emphasized the need for centralized log collection as necessary for success in AWS migration:
IT organizations suddenly gain access to thousands of servers running more applications than ever that also wind up generating more log files than most IT organizations can possibly cope with. … In order to be successful in the cloud, IT organizations need to relentlessly measure how both systems and applications are performing. The problem is that log collection and processing is painful.
All of the individual components Vogels described each have their own outputs of important log data that need to be analyzed. To ensure a successful AWS migration, logs from your applications, infrastructure and S3 Access buckets—and much more—all must be monitored to ensure all of the parts of your new AWS environment are operating in a way that is secure, highly available and scalable. (To help IT organizations, we have published a free e-book on AWS log analytics with the open source ELK Stack.)
AWS: A Final Note
AWS is the preferable public (IaaS) cloud choice for enterprises today, and it looks set to stay this way. To achieve a successful transition during a migration of workloads and products to AWS, the process needs to be carefully planned and implemented in a step-by-step fashion that will prove the benefits of the move to all of the stakeholders.
