How do you guarantee your Service Level Agreement (SLA)?
As an enterprise architect who has supported multiple software as a service (SaaS) offerings in recent years, the first question that I’m always asked relates to the SLA. Most businesses are interested in uptime, systems availability, network resiliency, fail-over options and capacity.As an example, a viable cloud provider will feel comfortable responding to the following concerns:
• Do you offer 24/7 support?
• How do you manage monitoring of the systems?
– Do you monitor all systems and applications?
– Do you monitor only uptime or do you include transactional monitoring?
– Would I receive alerts as well?
• How do you guarantee global connectivity and consistent speed?
– This presents an interesting component as most businesses now rely on global customer outreach. For example, an Italian customer should have the same digital experience as a customer in Taiwan.
• Do you have dual link or pairing connectivity to your data centers?
– In many cases, we are now seeing cloud providers that provide dual link with additional fail-over through their own network. A company like SoftLayer has made a name for itself using such setup.
• Do you meet or exceed industry standards for systems availability?
– Industry standard for systems availability hover around 99.8%
Who is responsible for the support of the application, systems and environment?
Responsibility for support will depend on the type of cloud services that you choose:• Software as a service (SaaS)
– Vendor bears responsibility for applications, data, run time, middleware, etc.
• Infrastructure as a service (IaaS)
– Business bears responsibility for applications, data, run time, middleware and operating system
• Platform as a service (PaaS)
– Business bears responsibility for applications and data
How do you manage backup, fail-over and disaster?
This is the million-dollar question. For most businesses, the ultimate fear of losing data can only be alleviated if a cloud provider can demonstrate through its Business Continuity Plan (BCP) that it has the policy and procedures in place to ensure a strong disaster recovery stance. There are many different techniques to ensure proper data backup, and some are more expensive than others. At minimum, a dual backup strategy such as backup-to-disk (local) and backup-to-tape (external) will guarantee that your data will be available following a disaster.What type of integration do you offer for connecting to third-party?
As the cloud evolves, more and more activities are performed through APIs among countless third-party providers. At times, such connectivity may require the creation of VPN tunnels or other esoteric connectivity options. It is important to know and understand what policies and options are offered as a major consideration for you to conduct business.How easily can I port my application to on-premise or another cloud provider?
As we all know, cloud is “pay-as-you-go/pay-as-you-grow.” You may find the need to move to another cloud provider or even decide to bring your business back home. This is a very important consideration as the ability to port your application whenever and wherever guarantees that you will always have a choice and that you—and you alone—are in control of your business.I hope I answered your questions and provided a sense of direction when engaging with a cloud provider. Feel free to contact me on twitter @thinkahead and let’s continue the discussion.