SLA for APIs: Monitoring Your API Service Level Agreement (SLA) 2017-09-11T10:04:10+00:00

SLA Measurement for APIs

The SLA Challenge

The goal of a Service Level Agreement (SLA) is to be able to agree with suppliers and customers that the quality of service you’re delivering or receiving meets an agreed standard.

The challenge is this: how do you mutually agree on SLAs and how do you decide if they are being met or not?

We’ll break this down for you, but it falls into the following issues:

  • Where do you measure from?
  • How do you measure up-time?
  • How do you understand consistency

These are complex questions, but we do have an answer.

LEARN MORE ABOUT CASC

Where do you measure from?

Location affects performance. Where the end-users of the API are, relative to the server handling the calls impacts performance a LOT more than people think. Worse. What a ping tool tells you about connection time, or your APM Tool tells you about server time might bear little or no relation to the experience of end users. Calls made in Asia can easily take twice as long as the same call made in the USA. Even inside the USA, there can be as much as a 25% difference in speed from coast to coast.

0%
Uptime
0ms
Latency (Total)
0ms
Ping Time
0%
Outliers

How do you count up-time?

In our survey of performance across over 1000 APIs we found there was a persistent error rate of about 0.5% across all APIs. The issue wasn’t necessarily the overall error rate but what that meant. All services have downtime for maintenance, for example. So the more important metric was how consistent the service was over time.

Finally, how do you take into account performance that falls outside normal performance expectations, we call them outliers, how do you factor them into your results?

Mutual Agreement

Critical to measuring an API is being able to mutually agree where and what is being measured and to have an accurate basis on which to set an SLA. Before you agree an SLA with a client you need to be able to confirm the following:

  • What is the basis of measuring the SLA?
  • Who is responsible for measurement?
  • How is adherence to the SLA to be reported?
  • What is the process for resolving SLA conflicts?

Finally, before you set an SLA, it’s usually a good idea to know what SLA you actually achieve before you sight up for it.

Mutual Agreement

Critical to measuring an API is being able to mutually agree where and what is being measured and to have an accurate basis on which to set an SLA. Before you agree an SLA with a client you need to be able to confirm the following:

  • What is the basis of measuring the SLA?
  • Who is responsible for measurement?
  • How is adherence to the SLA to be reported?
  • What is the process for resolving SLA conflicts?

Finally, before you set an SLA, it’s usually a good idea to know what SLA you actually achieve before you sight up for it.

Enter CASC

With Cloud API Service Consistency Scores, we go beyond any of the conventional metrics to give our clients a clear and easy to understand measure of quality and detailed regional performance measurement.

GET YOUR CASC SCORE