EXPERT SPEAK
Sachin Bhardwaj is Director of Marketing and Business Development at eHosting DataFort
Building a service level agreement with your hosting services provider
The details in a service level agreement depend on the likely duration of the relationship but should contain some basic terms elaborates Sachin Bhardwaj at eHosting DataFort.
With the increasing adoption of managed cloud services, the need for unambiguous, easy to apply and easy to follow service agreements are becoming more important. The level of effort that goes into preparation of a service level agreement between a services provider and the end-customer, is usually proportionate to the duration of the engagement.
A service level agreement is defined as an undertaking between a hosting services provider and an end-customer. This agreement describes various aspects of the service including quality, availability, and responsibilities. The service level agreement converts the hosting provider’ s services into a known function. This is particularly important where continuity and consistency of services from the hosting provider is of importance for the end-customer.
Another spin-off is the use of service level agreements by IT teams within an organisation to manage expectations of in-house stakeholders. This helps the IT team of the organisation to have its services justified, measured and benchmarked with other external IT service providers.
Hosting service providers need such agreements in order to manage customer expectations. They also need to define
the situations under which they are liable and not liable for deviations from the expected performance. For the end-customer having a prepared service level agreement allows the services of one provider to be objectively compared with another.
A well written and legally comprehensive service level agreement may include as many as two dozen different heads. Notwithstanding the number of clauses, a few of the more important ones appear below.
Services This is a detailed description of the services that will be provided and the end-customer groups that will receive specific services.
Availability This is a description of the uptime limits of a service. The service may also be described in terms of performance, business expectations and outcomes. The higher the availability requirement, the more expensive the service.
Downtime This is a description of the situation when the service is not available, exclusions for responsibility and force majeure, and the process of rebate linked to established failure.
Response time This is a description of the time to respond and rectify based on the nature of the failure or complaint.
Escalation For an externally hosted service, especially when there are time zone differences, this description is the most important for rectification of a fault. Going through a help desk or call centre repeatedly may not ensure timely rectification of the fault.
Notifications This describes the lead time required to be provided when there are any planned shutdowns or disruptions in the service. Planned maintenance cycles are usually known well in advance and can be built into business outcomes.
Benchmarks This is a description of the parameters used to gauge the upkeep of the service level agreement and the point of responsibility for producing, monitoring and escalating these benchmarks.
Help desk This is a description of how the help desk will respond based on the nature of the complaint or fault and the endcustomer flagging the complaint.
In conclusion, just because a service level agreement exists, it does not mean that the service provider and the end-customer will meet each other’ s expectations. Leveraging the terms in the service level agreement must be the means of last resort. There is much that can also be accomplished without resorting to it.
56 Issue 04 INTELLIGENT TECH CHANNELS