supabase / pg_net / 25989794116

93%

master: 93%

push

github

fix: flush pgstat counters from worker so autovacuum sees its writes

Background workers performing DML via SPI must call pgstat_report_stat()
themselves; regular user backends get this for free via tcop/postgres.c's
main loop. Without it, per-write counters (n_tup_ins, n_tup_del,
n_mod_since_analyze) for the worker's writes to net._http_response and
net.http_request_queue never reach shared stats. Autovacuum/autoanalyze
never see anything to vacuum, and net._http_response (written only by the
worker) silently bloats - eventually surfacing as 20-100s expiry DELETEs
on a bloated `created` index.

1 of 1 new or added line in 1 file covered. (100.0%)

498 of 533 relevant lines covered (93.43%)

204.25 hits per line