TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

A Mysterious PostgreSQL Performance Bug

14 pointsby wgynabout 4 years ago

1 comment

natmakaabout 4 years ago
When facing such ordeal and if you can accept to impact availability just VACUUM FREEZE the most impacted tables (or, if in doubt, the whole database).<p>If transaction activity is somewhat predictable adjust your autovacuum parameters per-table, using &quot;ALTER TABLE ((tablename)) SET&quot;, to have it kick not too early (constantly fiddling) and not too late (at worse leading to an automatic emergency DB locking preventing a wrapping). See <a href="https:&#x2F;&#x2F;www.postgresql.org&#x2F;docs&#x2F;current&#x2F;sql-altertable.html" rel="nofollow">https:&#x2F;&#x2F;www.postgresql.org&#x2F;docs&#x2F;current&#x2F;sql-altertable.html</a> Pertinent parameters: autovacuum_analyze_scale_factor autovacuum_analyze_threshold autovacuum_enabled = true autovacuum_vacuum_insert_scale_factor autovacuum_vacuum_insert_threshold autovacuum_vacuum_scale_factor autovacuum_vacuum_threshold autovacuum_vacuum_cost_delay autovacuum_vacuum_cost_limit<p>If multiple clients simultaneously hammer INSERT or UPDATE a table you can modify their sourcecode in order to have them cooperate by using PG&#x27;s advisory locking, see <a href="https:&#x2F;&#x2F;www.postgresql.org&#x2F;docs&#x2F;13&#x2F;explicit-locking.html#ADVISORY-LOCKS" rel="nofollow">https:&#x2F;&#x2F;www.postgresql.org&#x2F;docs&#x2F;13&#x2F;explicit-locking.html#ADV...</a>