what is the throughput in this usecase?

Question

Consider the system which gets 100 hits per second(HPS) and response time is 2 sec for each request. After initial 2 seconds, system is able to server 100 request per second which entered in to system 2 seconds before.

My question is can we say system has throughput of 100 just because 100 request are able to get out of the system. I believe not based on below definition.Per below definition if we are saying 100 throughput per second then those request should have been entered and exited within 1 second . But in above scenario 100 request entering into system are served 2 seconds later. So we can say 100 throughput per 2 seconds . Right ?

Throughput(QPS): This is the number of requests that are successfully executed/serviced per unit of time. For example, if the throughput is 50/minute, this means that on your server, per minute, 50 requests are executed successfully (accepted, processed and responded properly).

Answer 1

According to What is the Relationship Between Users and Hits Per Second? article:

Throughput is a measure of how many units of work are being processed. In the case of load testing, this is usually hits per second, also known as requests per second.

As per Apache JMeter Glossary:

Throughput is calculated as requests/unit of time. The time is calculated from the start of the first sample to the end of the last sample. This includes any intervals between samples, as it is supposed to represent the load on the server.

The formula is: Throughput = (number of requests) / (total time).

In your case if the system can process 100 requests in 2 seconds - its throughput is 100 requests in 2 seconds or 50 requests per second.

Answer 2

Where is your quote from ? it's confusing.

While this part is correct-

executed successfully (accepted, processed and responded properly)

it doesn't necessarily mean that it's the same request that was accepted processed and responded.

The quote doesn't differentiate between throughput, or rate, and latency.

Quoting from this link for example:

For OLTP workloads, performance can be broken down into throughput (often thought of as processing speed) and latency (time to achieve something).

We measure throughput as transactions per unit of time. In the case of the TPC-C results discussed below, we discuss transactions per minute (tpmC). Some benchmarks measure results as queries per second (QPS). This is a generic concept, and usually means a simple, non-transactional read or write command is sent to the server.

Latency is traditionally measured in milliseconds (ms) and will usually be distinguished as either read or write latencies depending upon the transaction.