Why can a hash table resize cause latency spikes, and how can you mitigate it?

Answer

Resizing often means allocating a bigger table and rehashing/copying all entries, which is O(n) work done at once. That can create a pause and a latency spike. Mitigations: pick a good load factor, pre-size when you know the size, use incremental rehashing (move entries gradually), or use a concurrent map implementation that spreads the work.