If you have ever run pg_ctl stop -m fast on a primary and watched it hang well past wal_sender_shutdown_timeout, you have met a bug that has been sitting in wal
pgvectorscale brings DiskANN's memory-efficient graph design to PostgreSQL, cutting vector search latency 9× versus HNSW when your index doesn't fit in RAM.
pgBackRest is unmaintained. Here's what to do next: an honest evaluation of Barman, WAL-G, and the realistic alternatives for production PostgreSQL backups.