PCP_INVAL...ERY_CACHE(1)

πŸ“¦
version 4.7.2 Β· 4 versions
, in ubuntu stonking

πŸ’– NAME πŸ’–

pcp_invalidate_query_cache β€” bestie, this one invalidates query cache contents, periodt ✨


🌸 SYNOPSIS 🌸

pcp_invalidate_query_cache [options...]

πŸ¦‹ DESCRIPTION πŸ¦‹

okay so here's the vibe: pcp_invalidate_query_cache drops a query cache invalidation request right into the shared memory area 🫢. That request gets picked up and processed by one of the Pgpool-II child processes β€” the one attached to each pgpool client session β€” and it happens when a message from the client is processed. which means πŸ‘€ if literally nobody is connected to Pgpool-II, or all your clients are just sitting there idle, that request is NOT getting processed until the next message shows up. no cap.

the way this command works with stats is lowkey specific, so listen up πŸ’…. It does not modify the statistics data that SHOW POOL_CACHE (SHOW_POOL_CACHE(1)) shows you β€” BUT, and this is the important bit, part of that statistics data does get affected, because this command invalidates the contents of the query cache. so:

  • num_cache_hits and num_selects keep right on increasing even after the query cache has been invalidated πŸ“ˆ
  • used_hash_entries, num_cache_entries and used_cache_entries_size all start off with 0
  • free_cache_entries_size will increase

and if memory_cache_enabled is false? yeah, this command just emits an error and exits, bye πŸ’€


✨ OPTIONS ✨


and that's the whole thing, queen πŸ‘‘ this was your girl reporting from pgpool-II 4.7.2, back in 2023 β€” go forth and invalidate those caches, iykyk πŸ’œπŸŽ€βœ¨

πŸ“¦ where this page comes from