Python · Mid-level · Theory

A nightly data processing script used to run for ten minutes, and now it takes two hours. How would you find out where it's spending its time?

~4 min by voice

How to approach it

This checks whether the candidate can systematically find the cause of a performance regression: measure with profilers first, tell CPU from I/O waiting, find the algorithmic problem and connect it to data growth or code changes, rather than optimizing at random.

Have you seen this question in a real interview?

Answers0

Avg. score—

Strong—

Practiced0

Topics

profiling performance

Similar questions

A service running on G1 periodically has pauses of one to two seconds, and it's hurting latency. What would you do, and when is it worth looking at a different garbage collector? Mid-level How is reflow different from repaint, and why does frequent reflow hurt performance? Junior A load test showed that at 500 requests per second, response time grew from 200 milliseconds to three seconds. How would you track down the cause, and what would you put in the report? Mid-level How do you track down a performance bottleneck in Python code? What's the most telling case from your own experience? Senior What happens in the browser between a change to the DOM or styles and pixels appearing on screen? How do you use that to keep animations and scrolling from janking? Senior

Answer by voice

The interviewer will ask this question and listen. Think out loud and ask for a hint if you need one. When you're done, pause or tap “That's my answer”. In about a minute, your score and the points you covered show up here.

Best with headphones · ~4 min