lock contention
Lock contention occurs when operations wait to acquire the same synchronization boundary.
Use this when Many independent operations queue behind one protected section.
without the term
“Adding more threads doesn't make it faster. They all seem to wait on the same lock around the cache. Can you speed it up without breaking thread safety?”
with the term · How do you ask a model for lock contention?
Reduce this measured lock contention. Preserve the protected invariant and compare correctness and throughput under concurrent load.
What is lock contention?
Many operations wait on the same lock, so throughput stops growing as you add load even though each operation is fast once it gets the lock.
Why does lock contention matter when vibe coding?
Removing a lock can improve a benchmark while introducing incorrect shared state.
How do you do lock contention?
Measure hold and wait time, then reduce unnecessary shared work without losing the invariant.
What goes wrong with lock contention?
Creating a per-key lock map that grows forever.