DEV Community

Saras Growth Space
Saras Growth Space

Posted on

Back-of-the-Envelope Estimation : Understanding QPS

In the previous post, we talked about what back-of-the-envelope estimation is and why it matters.

Now we move to the most important concept in system design:

QPS (Queries Per Second)


πŸš€ What is QPS?

QPS simply means:

How many requests your system handles every second

This is the starting point of almost every system design problem.


⏱️ The Key Conversion You Must Know

To convert daily traffic into real-time load:

  • 1 day β‰ˆ 100,000 seconds

This approximation makes calculations fast and simple.


πŸ”’ Basic Example

If your system gets:

  • 10 million requests per day

Then:

10,000,000 / 100,000 β‰ˆ 100 QPS
Enter fullscreen mode Exit fullscreen mode

So your system handles roughly:

100 requests per second


🧠 Think in Patterns (Not Calculations)

Instead of calculating every time, learn these patterns:

1M requests/day   β†’ ~10 QPS
10M requests/day  β†’ ~100 QPS
100M requests/day β†’ ~1,000 QPS
1B requests/day   β†’ ~10,000 QPS
Enter fullscreen mode Exit fullscreen mode

Once you internalize this, estimation becomes instant.


⚑ Why QPS Matters So Much

QPS directly impacts:

  • number of servers required
  • database load
  • system performance

Example:

  • 100 QPS β†’ small system
  • 10K QPS β†’ large system

🧩 Real Example

Let’s say:

  • 50 million requests/day

Using pattern:

  • 50M β‰ˆ 5 Γ— 10M
  • 10M β†’ 100 QPS

So:

  • 50M β†’ ~500 QPS

⚠️ Common Mistakes

❌ Trying to be too accurate

You don’t need exact numbers.

❌ Forgetting unit conversion

Always convert:

per day β†’ per second

❌ Getting stuck on math

Focus on approximation, not precision.


🧠 Key Insight

You’re not solving math problems.

You’re answering:

β€œIs this system handling hundreds, thousands, or millions of requests per second?”


πŸ‘‰ What’s next?

In the next post, we’ll cover:

  • Storage estimation
  • How systems grow over time
  • GB/day β†’ TB/year thinking

If this helped, drop your thoughts πŸ‘‡

Top comments (0)