debugging-strategies

Intermittent bugs require extensive logging of timing, state transitions, and external interactions. Look for race conditions in concurrent code, check async ordering, and stress test by running the scenario many times under varied load and timing.

Use git bisect to binary search commit history. Mark the current commit as bad and a known working commit as good, then test each checked-out midpoint, marking it good or bad until git identifies the exact faulty commit.

Use cProfile with pstats to profile Python code and identify the slowest functions by cumulative time. For interactive debugging, use breakpoint() or ipdb, and add structured logging to trace execution flow.

Monitor process.memoryUsage() for growing heap usage and generate heap snapshots with v8.writeHeapSnapshot(). In Chrome DevTools, compare heap snapshots before and after actions to find retained objects that should have been freed.

Use a debugger when you need to inspect complex state, step through execution, or set conditional breakpoints without modifying code. Console logging works for quick traces but becomes unwieldy for deep call stacks or timing-sensitive issues.