| From: | shihao zhong <zhong950419(at)gmail(dot)com> |
|---|---|
| To: | Bharath Rupireddy <bharath(dot)rupireddyforpostgres(at)gmail(dot)com> |
| Cc: | PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org>, Amit Kapila <amit(dot)kapila16(at)gmail(dot)com> |
| Subject: | Re: Parallel vacuum: wrong error context when the leader vacuums an index |
| Date: | 2026-09-29 06:11:28 |
| Message-ID: | CAGRkXqSj4=HUczJ8=OW2FLnkb5iRfYd8NTNS1OizGpcutCOpFA@mail.gmail.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Hi Bharath,
+1 for the patch. The fix looks right to me.
The new test fails on master with
CONTEXT: while scanning relation "public.nbtree_vacuum_error_context"
and passes with the patch. The same holds for an error in the cleanup
pass, an index pass in the middle of the heap scan (master says "while
scanning block 1 of relation"), and an error relayed from a worker. With
the patch all of them name the index, and the worker error no longer ends
with the stale "while scanning relation" line.
One Nit:
The test only covers bulk delete. The cleanup pass reports "while
cleaning up index" and is not tested. A BRIN index on an expression that
raises an error hits it, since BRIN cleanup summarizes the new ranges.
Thanks,
Shihao
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Min, Baohong | 2026-09-29 06:11:49 | RE: [PATCH] Reduce LWLockWaitListLock() cache-line contention with adaptive spin reads |
| Previous Message | Chao Li | 2026-09-29 06:09:40 | pg_resetwal: Fix handling of commit timestamp XIDs |