YOUR BRIEF
What to work through
Work through a job scheduler as a spoken system design interview. Start by clarifying the scope, sketch a working design and explain the trade-offs behind your choices.
- Schedule a job to run now, at a future time, or on a recurring schedule such as every day at 10:00
- See the status of each job and its runs
- Retry failed runs
- Priorities, and jobs that depend on other jobs
BEFORE YOU START
Constraints to discuss
- Runs start within 2 seconds of their scheduled time
- 10,000 job runs a second
- Every run happens at least once
- One tenant’s flood of jobs cannot starve the others
Need a starting point?
Watch for these common gaps as you prepare:
- Cron on a single machine, a single point of failure
- Polling the whole jobs table every second
- No lease, so a dead worker’s job is lost
- Retrying tasks that are not idempotent
- No plan for runs missed during an outage
Why this question is in the collection
These sources connect the problem to company interview reports or preparation guides. They do not guarantee what a company will ask you.
- OpenAI: PracHub lists a text-to-video GPU job scheduler with one GPU per task and fair queueing as an OpenAI question, posted in September 2026.
- Google: PracHub lists a scheduler for long-running GPU jobs with tenant fairness as a Google technical screen question, posted in June 2026.
- Meta: Aced shows a Meta senior (L5) candidate reporting a distributed job scheduling question.