Interactive explainers of backend and systems internals — the casing removed.https://cutaway.gilla.fun/enhttps://cutaway.gilla.fun/lsm-tree-compaction/https://cutaway.gilla.fun/lsm-tree-compaction/Your B-tree ingest tops out at 5k rows/s and the SSD is bored — random writes are the bottleneck, not bandwidth. An LSM tree fixes the write path by making reads do the work later. Pile up uncompacted files and watch read amplification climb into the danger zone.Thu, 11 Jun 2026 00:00:00 GMThttps://cutaway.gilla.fun/pgbouncer-pool-modes/https://cutaway.gilla.fun/pgbouncer-pool-modes/A Postgres backend is a whole process, so a few hundred of them is a real ceiling. A pooler multiplexes thousands of clients onto a handful of server connections — and the mode you pick decides which of your queries silently break.Thu, 11 Jun 2026 00:00:00 GMThttps://cutaway.gilla.fun/raft-leader-election/https://cutaway.gilla.fun/raft-leader-election/A primary on the wrong side of a partition is still taking writes. The fix is not a faster failover detector — it is a rule about who is allowed to win. Cut the links yourself and watch Raft refuse to elect two leaders.Thu, 11 Jun 2026 00:00:00 GMThttps://cutaway.gilla.fun/temporal-deterministic-replay/https://cutaway.gilla.fun/temporal-deterministic-replay/A worker is OOM-killed 40 minutes into an order workflow. It restarts and the order finishes as if nothing happened — nobody wrote recovery code. The trick is replaying recorded history through deterministic code. Crash the worker yourself and watch the local variables refill.Thu, 11 Jun 2026 00:00:00 GMThttps://cutaway.gilla.fun/wal-crash-recovery/https://cutaway.gilla.fun/wal-crash-recovery/COMMIT returned OK, then the box lost power. Whether that row is still there comes down to one fsync. Kill the database mid-write and watch recovery replay the log.Thu, 11 Jun 2026 00:00:00 GMT