public inbox for [email protected]
help / color / mirror / Atom feedFrom: SATYANARAYANA NARLAPURAM <[email protected]>
To: Daniil Davydov <[email protected]>
Cc: Bharath Rupireddy <[email protected]>
Cc: Masahiko Sawada <[email protected]>
Cc: Sami Imseih <[email protected]>
Cc: Alexander Korotkov <[email protected]>
Cc: Matheus Alcantara <[email protected]>
Cc: Maxim Orlov <[email protected]>
Cc: Postgres hackers <[email protected]>
Subject: Re: POC: Parallel processing of indexes in autovacuum
Date: Mon, 30 Mar 2026 17:14:39 -0700
Message-ID: <CAHg+QDdRyxC-cBk7CK-=pnfqFNWh6BeFDDnz3CSUPyoTbdUJ+A@mail.gmail.com> (raw)
In-Reply-To: <CAJDiXgi73x7h0=UoXriFjskRB6htZ-uqXKqvWN3RefuxbP93gA@mail.gmail.com>
References: <CACG=ezZOrNsuLoETLD1gAswZMuH2nGGq7Ogcc0QOE5hhWaw=cw@mail.gmail.com>
<CAD21AoCdx5ZNS_cO7bYz1Zfb+Kw1kuJV2wtewrz7T1pPpjcWGw@mail.gmail.com>
<CAJDiXgi6ZQOoSEqj9RyZMEh+HHBtmW0+PHD85UNPtKch8ubvdg@mail.gmail.com>
<CAD21AoBcoA-i-pJ_=y+jg14R8_QaJA1iwktCnu5i-C=yXDFPdA@mail.gmail.com>
<CAJDiXgjnUdE6Sk4M0unmT+9dULyFAxcum2txQKpWTuo4uQ_oXQ@mail.gmail.com>
<CAD21AoBTZdVR93JBo620B=MX-K8cdm3VRbjrBr_Vcpngk3AjVw@mail.gmail.com>
<CAA5RZ0vfBg=c_0Sa1Tpxv8tueeBk8C5qTf9TrxKBbXUqPc99Ag@mail.gmail.com>
<CAD21AoBgvUeWS8ZsXBahA1XdYayK6DJ6dx49d6Xpii-iH+Hrwg@mail.gmail.com>
<CAA5RZ0vF+Lr-jU1LAZWTGUjboUETk8oLvaNBbA5ozX6dau+how@mail.gmail.com>
<CAJDiXggueLSGMNRmLshbmFRfbo4jzks0W8bLDfUSRZ-61fPVEQ@mail.gmail.com>
<CAFY6G8cJ=DRTX75pOGerH6sk39dRt+7MSH+y_qppDdhPs=qdQA@mail.gmail.com>
<CAJDiXgg1t6wk9NjyMUTm1iKqM9GtdQ_wrEchBtz3xjWBZM8W8A@mail.gmail.com>
<CAD21AoAC0=Xi38RQcAO4A+vdmoXToZMoHfbS=KLT49fAOTH_gA@mail.gmail.com>
<CAJDiXgiD+AZKhJSn-FSRVQxtDLmJd95wDu4wtKniQF5==1JcjQ@mail.gmail.com>
<CAD21AoAM8KsqNhrZYJuf7odvxcTC0TumXazJc-r_wC5KnDFDPg@mail.gmail.com>
<CAJDiXghbcOC9OOj3ampxuyqXH0geggnosnrYUHGygkpss-RtxA@mail.gmail.com>
<CAD21AoAPnq0vrcGgeN++r1GoL8Kza7jaGL=TNzuBn6+MkR=rUQ@mail.gmail.com>
<CAJDiXghmsbTmnm--9B5bbuZXa1OL7SZ0HYppX3tx9XsdwfJBhA@mail.gmail.com>
<[email protected]>
<CAJDiXgiYiX+azuR76DcVx8fZn57m_4v6cB14-GW34mWa=qudFQ@mail.gmail.com>
<CAD21AoDtPpkkQ_h1yf4oTx1qn4SRdTeVY3qs+9J07fYqa_4Gww@mail.gmail.com>
<CAJDiXgi7KB7wSQ=Ux=ngdaCvJnJ5x-ehvTyiuZez+5uKHtV6iQ@mail.gmail.com>
<CAD21AoCcHKKXsr9Oh736ejckqqS1i430xGEyJ=JP5OL0ExyP1A@mail.gmail.com>
<CAJDiXghaFT_1sSv3q8mjyZ_RLZDgiogg0mWRvLxSWvkUi2CcLg@mail.gmail.com>
<CAA5RZ0u63W41OmcEO+HLs4CSo-Sd3J+Q-4=04iud8V=xX4iUrA@mail.gmail.com>
<CAJDiXgin1TXniVGJKzOTA=F9K342uVfm6O0EmubTVB=F+XSrbA@mail.gmail.com>
<CAD21AoDadzAwibxf-+urjx=XL+eVu8=Ut-Lh2GxXUt32LbPG3Q@mail.gmail.com>
<CAD21AoD6HhraqhOgkQJOrr0ixZkAZuqJRpzGv-B+_-ad6d5aPw@mail.gmail.com>
<CAJDiXgiGSpqMQSOx-cVO_LtcB5GWHBy9ph7oOR4ebbX8A==kgw@mail.gmail.com>
<CAD21AoBRRXbNJEvCjS-0XZgCEeRBzQPKmrSDjJ3wZ8TN28vaCQ@mail.gmail.com>
<CAPpHfduBJfMcojvmYHUo8b_C=0cxRy1N+tNiNGoA3RAZq2ApaA@mail.gmail.com>
<CAD21AoC82NeHKXc965pPUZO2eyo1U7P6cmfRJbrcPDcnd7_6hw@mail.gmail.com>
<CAJDiXghP2kXnEz+cj3rAWNM3NdKSB_4WtnngFXpVz2omPhGr5A@mail.gmail.com>
<CAD21AoA0bnRZC_OqKMnH-Ln+OZ9z9k56j2c_MXj8pw69O-wkBw@mail.gmail.com>
<CAA5RZ0sSXDza7_nUUbhHL_Sws+M+HR1daKJPXHpdLuNCkwUgUg@mail.gmail.com>
<CAJDiXggrBsbzOisf+Nu8pZkYGrpUZaFbosL1Wbm3kKxzTm4xgw@mail.gmail.com>
<CAA5RZ0tbiPcgQEjnhdnjz6qSjfRsGrr8jGCaMcrMaoPpax3wig@mail.gmail.com>
<CAJDiXgjt5ZmK2uvS0E8Ztt5ePYmq8Ze_dG05Zo2NUsKLHCEuYA@mail.gmail.com>
<CAD21AoB7v5tLPXLK=qmtt6PaEC1f+Fb-gh+MwAbXfm6x4eZGNw@mail.gmail.com>
<CAJDiXghwtUbiFnAh3nSaxTk8KFupQuMbp+g4z3wOLoQfMuqgDg@mail.gmail.com>
<CAJDiXgjoNd4BF19HNY_FAcDUqiqsfw8cGhNOJwBxahB8P38E3Q@mail.gmail.com>
<CAD21AoBT1LWqPZkcHpVMVh0ZOXUneO=p61t0i8cQ+kOP9qfODQ@mail.gmail.com>
<CAJDiXggL=J0nV7PfBsMW9+UOU3KUp1jNBM9Gov1JvAX7aG_U1g@mail.gmail.com>
<CAD21AoDz-1Zf9DOJJrdcB2=eNA4UdywthkowNp_dHmOGC-yV_g@mail.gmail.com>
<CAJDiXgjzphJ313=aDwbvryHpmTi6AqE+-5crysTtzKv01-vkzA@mail.gmail.com>
<CAD21AoD7_4gsQ2a82zO3SaRwjdw_3tyiYDHNFPUKQ5DAA5HOtA@mail.gmail.com>
<CAJDiXggY1QzNde6_HhpzneLc9dYqmWZ+PY39cuBXYdcCTuoJBA@mail.gmail.com>
<CAD21AoCFPiS2jcMA1JaV1kT8xrGz5BpN7iBP_gCgRuaANEbciA@mail.gmail.com>
<CAJDiXgh6jmNGR3uOB_6YeGhNkR2=HdTdEYjmHXdumNzyY4MckQ@mail.gmail.com>
<CAD21AoDs3SOXeAEoCRizfEKybpRkE7t7poX0+iZ6MM1MFWMsfA@mail.gmail.com>
<CAJDiXgjTkuqSPerC_nasxDz6d2Komf1ipYKV6SupDRnc9yhO9w@mail.gmail.com>
<CAD21AoAXMjX03h5K84u0heBLU+fqGgWBGBDwnBDGSs=DhyF9pQ@mail.gmail.com>
<CAJDiXghjZEAYboGhujgGvY9=RiFD01ERHVVF+NQMuuAKVZDmDQ@mail.gmail.com>
<CAD21AoAD+N5SxBr0qL7TeWnvq4iYmFT=DyWdNLQPB-XntYkwEg@mail.gmail.com>
<CAJDiXgjgn87sH1-MmONPKkeYJG83C0ChrYkYn9UcRonLhOOfOw@mail.gmail.com>
<CAD21AoCoJYauWO78M-CGdHpYfcqEZVV5a1Z-7wWB=-G-x8EVFg@mail.gmail.com>
<CAJDiXghaazbrQMZZS08d9Ffh2y4w05TgH9dpBhqChv1qNTp+xA@mail.gmail.com>
<CAD21AoDbaNtLrFRxG9OG5WrBd7DCs4q+CfJd8AJTBEqRri4WeQ@mail.gmail.com>
<CAJDiXgjjd1jL86B--AyRo2tDM1Wiu+7Pduwh5d0u_UM8GRugvw@mail.gmail.com>
<CAD21AoBo_wS7y0X7_7ajEFkptzo9ZrF8RFNRnu2Xe8XL74o0SQ@mail.gmail.com>
<CAJDiXggH1bW=4n+55CGLvs_sRU4SYNXwYLZ37wvJ5H_3yURSPw@mail.gmail.com>
<CAD21AoDxhN8Z6Lx1ZicBXKkbMsRQqEXiq4ALs4uaD648iSvXoA@mail.gmail.com>
<CAJDiXgh3Dg2f5k3xRJnzoY39jQENUhh125ArYapXkSu5D7JJuw@mail.gmail.com>
<CAD21AoBYc7L7W4dRdxeoJzOH5OgpiCAtKz-54iX4Ufn8PnQoww@mail.gmail.com>
<CAJDiXgi8X-DMb92v5WHLCNxDHxH9gO8WQxOMtdpmU7X=WXCiuQ@mail.gmail.com>
<CAD21AoDKxs0UrTwa3rkP+kE9AzccabpK7G-Tk=HYneaFTZBtiA@mail.gmail.com>
<CALj2ACUJ0TtYWtFuXXVf0aLES8tfZePXnB8WQ=0KCrNaABzQVg@mail.gmail.com>
<CAJDiXgj=-R1z7H7+npm-o+q6YBkr5_6Qe=1wcy47ovAqej4TkA@mail.gmail.com>
<CAHg+QDehxaJEd1Yp1MpW8UO71xmbasy7t2GZGvqOYwkr0md8DQ@mail.gmail.com>
<CAJDiXgi73x7h0=UoXriFjskRB6htZ-uqXKqvWN3RefuxbP93gA@mail.gmail.com>
Hi
On Mon, Mar 30, 2026 at 1:44 AM Daniil Davydov <[email protected]> wrote:
> Hi,
>
> On Mon, Mar 30, 2026 at 7:17 AM SATYANARAYANA NARLAPURAM
> <[email protected]> wrote:
> >
> > Thank you for working on this, very useful feature. Sharing a few
> thoughts:
> >
> > 1. Shouldn't we also cap by max_parallel_workers to avoid wasting DSM
> resources in parallel_vacuum_compute_workers?
>
> Actually, autovacuum_max_parallel_workers is already limited by
> max_parallel_workers. It is not clear for me why we allow setting this GUC
> higher than max_parallel_workers, but if this happens, I think it is a
> user's
> misconfiguration.
>
> > 2. Is it intentional that other autovacuum workers not yield cost limits
> to the parallel auto vacuum workers? Cost limits are distributed first
> equally to the autovacuum workers.
> > and then they share that. Therefore, parallel workers will be heavily
> throttled. IIUC, this problem doesn't exist with manual vacuum.
> > If we don't fix this, at least we should document this.
>
> Parallel a/v workers inherit cost based parameters (including the
> vacuum_cost_limit) from the leader worker. Do you mean that this can be too
> low value for parallel operation? If so, user can manually increase the
> vacuum_cost_limit reloption for those tables, where parallel a/v sleeps too
> much (due to cost delay).
>
> BTW, describing the cost limit propagation to the parallel a/v workers is
> worth mentioning in the documentation. I'll add it in the next patch
> version.
>
> > 3. Additionally, is there a point where, based on the cost limits,
> launching additional workers becomes counterproductive compared to running
> fewer workers and preventing it?
>
> I don't think that we can possibly find a universal limit that will be
> appropriate for all possible configurations. By now we are using a pretty
> simple formula for parallel degree calculation. Since user have several
> ways
> to affect this formula, I guess that there will be no problems with it
> (except
> my concerns about opt-out style).
>
> > 4. Would it make sense to add a table level override to disable
> parallelism or set parallel worker count?
>
> We already have the "autovacuum_parallel_workers" reloption that is used as
> an additional limit for the number of parallel workers. In particular, this
> reloption can be used to disable parallelism at all.
>
> >
> > I ran some perf tests to show the improvements with parallel vacuum and
> shared below.
>
> Thank you very much!
>
> > Observations:
> >
> > 1. Parallel autovacuum provides consistent speedup. With cost_limit=200
> and
> > 7 workers, vacuum completes 1.41x faster (71s -> 50s). With
> cost_limit=60,
> > the speedup is 1.25x (194s -> 154s).
> > 2. I see the benefit comes from parallelizing index vacuum. With 8
> indexes totaling
> > ~530 MB, parallel workers scan indexes concurrently instead of the
> leader
> > scanning them one by one. The leader's CPU user time drops from ~3s to
> > ~0.8s as index work is offloaded
> >
>
> 1.41 speedup with 7 parallel workers may not seem like a great win, but it
> is
> a whole time of autovacuum operation (not only index bulkdel/cleanup) with
> pretty small indexes.
>
> May I ask you to run the same test with a higher table's size (several
> dozen
> gigabytes)? I think the results will be more "expressive".
>
I ran it with a Billion rows in a table with 8 indexes. The improvement
with 7 workers is 1.8x.
Please note that there is a fixed overhead in other vacuum steps, for
example heap scan.
In the environments where cost-based delay is used (the default), benefits
will be modest
unless vacuum_cost_delay is set to sufficiently large value.
*Hardware:* CPU: Intel Xeon Platinum 8573C, 1 socket × 8 cores × 2
threads = 16 vCPUs
RAM: 128 GB (131,900 MB)
Swap: None
*Workload Description*
*Table Schema:*
CREATE TABLE avtest (
id bigint PRIMARY KEY,
col1 int, -- random()*1e9
col2 int, -- random()*1e9
col3 int, -- random()*1e9
col4 int, -- random()*1e9
col5 int, -- random()*1e9
col6 text, -- 'text_' || random()*1e6 (short text ~10
chars)
col7 timestamp, -- now() - random()*365 days
padding text -- repeat('x', 50)
) WITH (fillfactor = 90);
*Indexes (8 total):*
avtest_pkey — btree on (id) bigint
idx_av_col1 — btree on (col1) int
idx_av_col2 — btree on (col2) int
idx_av_col3 — btree on (col3) int
idx_av_col4 — btree on (col4) int
idx_av_col5 — btree on (col5) int
idx_av_col6 — btree on (col6) text
idx_av_col7 — btree on (col7) timestamp
Dead Tuple Generation:
DELETE FROM avtest WHERE id % 5 IN (1, 2);
This deletes exactly 40% of rows, uniformly distributed across all pages.
Vacuum Trigger:
Autovacuum is triggered naturally by lowering the threshold to 0 and
setting
scale_factor to a value that causes immediate launch after the DELETE.
Worker Configurations Tested:
0 workers — leader-only vacuum (baseline, no parallelism)
2 workers — leader + 2 parallel workers (3 processes total)
4 workers — leader + 4 parallel workers (5 processes total)
7 workers — leader + 7 parallel workers (8 processes total, 1 per index)
Dataset:
Rows: 1,000,000,000
Heap size: 139 GB
Total size: 279 GB (heap + 8 indexes)
Dead tuples: 400,000,000 (40%)
Index Sizes:
avtest_pkey 21 GB (bigint)
idx_av_col7 21 GB (timestamp)
idx_av_col1 18 GB (int)
idx_av_col2 18 GB (int)
idx_av_col3 18 GB (int)
idx_av_col4 18 GB (int)
idx_av_col5 18 GB (int)
idx_av_col6 7 GB (text — shorter keys, smaller index)
Total indexes: 139 GB
Server Settings:
shared_buffers = 96GB
maintenance_work_mem = 1GB
max_wal_size = 100GB
checkpoint_timeout = 1h
autovacuum_vacuum_cost_delay = 0ms (NO throttling)
autovacuum_vacuum_cost_limit = 1000
*Summary:*
Workers Avg(s) Min(s) Max(s) Speedup Time Saved
------- ------ ------ ------ ------- ----------
0 1645.93 1645.01 1646.84 1.00x —
2 1276.35 1275.64 1277.05 1.29x 369.58s (6.2 min)
4 1052.62 1048.92 1056.32 1.56x 593.31s (9.9 min)
7 892.23 886.59 897.86 1.84x 753.70s (12.6 min)
Thanks,
Satya
view thread (112+ messages) latest in thread
reply
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Reply to all the recipients using the --to and --cc options:
reply via email
To: [email protected]
Cc: [email protected], [email protected], [email protected], [email protected], [email protected], [email protected], [email protected], [email protected], [email protected]
Subject: Re: POC: Parallel processing of indexes in autovacuum
In-Reply-To: <CAHg+QDdRyxC-cBk7CK-=pnfqFNWh6BeFDDnz3CSUPyoTbdUJ+A@mail.gmail.com>
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
This inbox is served by agora; see mirroring instructions
for how to clone and mirror all data and code used for this inbox