hard20 pts · first answer countspostgresqlperformance

WAL volume spikes after checkpoints

Logs show "checkpoints are occurring too frequently" and WAL volume is very high right after each checkpoint. What explains it and what helps?

A.The first change to each page after a checkpoint writes a full page image; raising max_wal_size (and checkpoint_timeout) means fewer checkpoints and fewer full page imagesB.Checkpoints delete WAL, so the volume only looks high; no change is neededC.Autovacuum is disabled; enable it to reduce WALD.shared_buffers is too large; reduce it to 128 MB

← Primary demoted when etcd is unreachable What encryption at rest protects →