From: | "houzj(dot)fnst(at)fujitsu(dot)com" <houzj(dot)fnst(at)fujitsu(dot)com> |
---|---|
To: | Masahiko Sawada <sawada(dot)mshk(at)gmail(dot)com> |
Cc: | PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org>, Peter Geoghegan <pg(at)bowt(dot)ie>, Andres Freund <andres(at)anarazel(dot)de>, Amit Kapila <amit(dot)kapila16(at)gmail(dot)com> |
Subject: | RE: parallel vacuum comments |
Date: | 2021-12-08 03:22:16 |
Message-ID: | OS0PR01MB5716B6C7A90AC112E0240AD9946F9@OS0PR01MB5716.jpnprd01.prod.outlook.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On Tuesday, December 7, 2021 1:42 PM Masahiko Sawada <sawada(dot)mshk(at)gmail(dot)com> wrote:
> I've attached an updated patch. I've removed 0003 patch that added
> regression tests as per discussion. Regarding the terminology like "bulkdel"
> and "cleanup" you pointed out, I've done that in 0002 patch while moving the
> code to vacuumparallel.c. In that file, we can consistently use the terms
> "bulkdel" and "cleanup" instead of "vacuum"
> and "cleanup".
Hi,
Thanks for updating the patch.
I noticed few minor things.
0001
1)
* Skip processing indexes that are unsafe for workers (these are
- * processed in do_serial_processing_for_unsafe_indexes() by leader)
+ * processed in parallel_vacuum_process_unsafe_indexes() by leader)
It might be clearer to mention that the index to be skipped are unsafe OR not
worthwhile.
2)
+ /* Set index vacuum status and mark as parallel safe or not */
+ for (int i = 0; i < pvc->nindexes; i++)
+ {
...
+ pindstats->parallel_workers_can_process =
+ parallel_vacuum_index_is_parallel_safe(vacrel,
+ vacrel->indrels[i],
+ vacuum);
For the comments above the loop, maybe better to mention we are marking whether
worker can process the index(not only safe/unsafe).
0002
3)
+ /*
+ * Skip indexes that are unsuitable target for parallel index vacuum
+ */
+ if (parallel_vacuum_should_skip_index(indrel))
+ continue;
+
It seems we can use will_parallel_vacuum[] here instead of invoking the function
again.
Best regards,
Hou zj
From | Date | Subject | |
---|---|---|---|
Next Message | Tomas Vondra | 2021-12-08 03:26:28 | Re: 回复:Re: Is it worth pushing conditions to sublink/subplan? |
Previous Message | Tomas Vondra | 2021-12-08 02:22:10 | Re: Is there a way (except from server logs) to know the kind of on-going/last checkpoint? |