When start-ups and enterprises first evaluate cloud providers, they often choose an out of the box solution that fits their immediate needs. Their hosting provider promises great customer service, but they have essentially the same stack as thousands of other customers and still need a large SysOps staff to monitor their infrastructure.
This solution may be adequate for several months or even years. But most enterprises ultimately find that a one-size-fits-all cloud solution no longer actually fits – if it ever did in the first place. Whether due to performance hiccups, rapid growth, a desire for a more automated DevOps approach, or new compliance challenges, they begin looking for other options.
The combination of a powerful infrastructure like AWS and a hands-on managed service provider is frequently the best solution.
A public cloud like AWS is a set of expertly designed tools – powerful, revolutionary, capable of building an incredible machine. AWS does not, however, tell you how to architect an environment from the resources it provides. It is an unassembled F-22 without a pilot or mechanic. As complexity increases, enterprises need someone to assist them beyond just fixing things when they break. They need someone to customise, right-size, and automate their infrastructure for their specific requirements and integrate with their internal IT teams – not just answer the phone when they call. They need someone who will take responsibility for security concerns and have the capacity to provide detailed reports to meet audit requirements.
In fact, Amazon understands that infrastructure is only half the battle for most enterprises. That is why Amazon invites Consulting Partners to help enterprises understand the full possibilities of the cloud. A managed cloud services provider allows enterprise clients to take full advantage of AWS services beyond elastic computing and VM provisioning.
These are the functions and services enterprises are usually looking for in a managed services partner:
1. Cloud migration
Most enterprises need help learning about which AWS resources can best replicate or improve their bare metal infrastructure. While AWS has provided extensive documentation for its services, each application is unique and internal IT teams often do not have the time to perform extensive testing on instance size/capacity, security, etc. The best managed service partners will offer a thorough audit and discovery process on the current environment, not a cursory “lift and drop” solution that usually results in improperly-sized instances and poor performance. It may take time to test the application, beginning with the smallest possible instance and conducting extensive performance testing to ensure the solution is cost-effective while maintaining the level of performance they expect from on-premise infrastructure.
A key part of cloud migration is also expertise in both cloud engineering and traditional IT. If your managed service provider was “born in the cloud,” meaning that they opened their doors five or so years ago and employ only cloud engineers, will they understand an enterprise’s legacy applications? Will they understand why the database in on-premises infrastructure needs a special blend of resources in the public cloud or how to get higher I/O out of AWS? Are they capable of understanding when an application needs to be hosted on a private cloud? Only a partner with extensive managed private hosting experience and AWS expertise can understand where an enterprise is now help get them where they are going.
2. Cloud management, not customer service or consulting
Obviously, managing cloud infrastructure requires a very different set of skills, and maintaining an external team with cloud expertise is often more cost-effective than maintaining a staff of cloud engineers in-house (or suffering from downtime due to a lack of staff experience). There are, however, vastly different levels of support; some offer ticket support, others offer 24/7/365 phone support, and still others integrate with your internal IT team to support code pushes, seasonal events, etc. Enterprises are generally looking for the latter, no matter how “fanatical” the phone support promises to be.
Some enterprises get stuck with a managed service provider that only really offers consulting services. While they may provide some technology guidance and implementation, they hold no responsibility for the ultimate product and must be hired again if the infrastructure ever changes. Before you hire a managed service provider, understand where the responsibility ends and who is responsible for what. As we explain in detail here, it is often not a long-term cost savings to engage a consulting partner rather than a managed services partner.
3. Cloud SLAs
AWS offers a high guarantee on uptime. But for mission-critical IT applications, this guarantee may not be enough. Managed Services providers that offer 100% uptime SLAs do so because they are able to configure a unique blend of native AWS and third party tools to create a self-healing, auto scaling environment that never goes down. Very few providers are able offer this, due to the fact that they must constantly monitor and test the environment to meet this requirement. Smaller cloud providers are certainly unable to guarantee zero downtime.
DevOps is a buzzword that has been used to apply to nearly any development framework. A true DevOps shop will encourage clients to focus on automation and integration; they will make it possible for a client to bring up environments in new regions in a matter of hours and use a configuration management tool like Puppet to maintain a single source of consistent, documented system configuration, deploy environments prescriptively, and enforce a mature Software Development Lifecycle. An automated AWS environment is neither easy nor automatic, and does require a significant upfront investment to bake AMIs, write custom configuration management scripts, etc. to be able to deploy new environment in a matter of hours.
5. Security services
After a string of high profile security attacks, security is the #1 concern of CIOs and CTOs in 2015. While Amazon guarantees the security of the physical infrastructure with the kind of physical security measures that few datacenters can boast, the user is responsible for security “in” the cloud. Enterprises look to cloud managed service providers to minimize risk and carry a significant portion of the responsibility for infrastructure security.
AWS has a number of native resources that make enterprise-grade security possible.
6. Compliance services
Beyond traditional security best practices, compliance requirements necessitate a level of monitoring, reporting, and data storage that internal IT teams are either unfamiliar with or do not know how to translate to the cloud – exposing a multi-million dollar fine risk for large enterprises. Enterprises will often need to sign a BAA with AWS and their managed services provider. AWS has in fact Amazon has always been on the leading edge of compliant storage solutions in the cloud.
Managed service providers that specialize in compliance, have a long history managing compliant infrastructure, and have been through multiple audits are better equipped to deal with security threats – even if the specific application does not have compliance requirements. Providers that specialise in compliance use security best practices as default for all instances.
7. Cloud integration
Increasingly, enterprises have realised that enlisting the help of multiple service providers for their infrastructure can lead to fragmentation, poor communication, excessive contract negotiation work, and wasted resources. They need a managed service provider who will be able to move them to AWS while still hosting some legacy applications in a private cloud that is connected to their AWS deployment and united under a single SLA and SOW. This facilitates the migration over time of applications to AWS.
As the above list outlines, infrastructure built a managed services provider and powered by AWS is much more sophisticated than customer service on top of a smaller, less frequently updated cloud. With the right managed services provider, it is possible to create an enterprise-grade, self-healing, dedicated and highly secure infrastructure on AWS that has significant advantages over other solutions.