In the last chapter, I tried to introduce you to the topic of service level agreement (SLA). An SLA is a contract that includes a guaranteed value for the availability. In addition to this guaranteed value, so-called service level objectives (SLOs) can also be used. But SLOs are not guarantees, but goals in the performance area. In other words, the service is just trying to reach a certain level. The entire environment and interaction with other services can have a negative impact.

Typical SLOs can be found in the following areas:

• Throughput: How many operations can be performed in a given period of time?
• Concurrency: How many operations can be performed simultaneously?
• Latency: How long does the system take to perform an operation?
• Headroom: How much capacity does the system have to allow growth?
• Error rate: How many exceptions does the system generate while performing operations?

Now we have learned a lot of theory about performance. But we also want to work in this area. For example, we want to know: where does a bottleneck exist in my application?