Hacker News Reader

I'd guess using uuids for the keys instead of ROWIDs might help. I'd also guess that UNIQUE constraints must also trigger a conflict and that "close together" in index btrees also does. If that's true then using a uuid primary key is defeated by the indices that you'd use to work around it.

View on HN

Using unordered UUIDs for keys will also slow down inserts since the b-tree needs rebalancing so often.

A uuid key would crush performance vs rowid, though, especially for ordered scans, and you’d still have conflicts if parallel r/w accesses are temporally correlated (you wrote X and Y at the ~same time, and tend to read them at the ~same time)

Reply on news.ycombinator.com