Date: Wed, 05 Apr 2023 12:37:41 +0530
From: Karthik Krishnakumar <karthikk@zohocorp.com>
To: "pgsql-admin" <pgsql-admin@lists.postgresql.org>
Message-Id: <18750398329.1002ca2ba316.8870907212977424118@zohocorp.com>
In-Reply-To: 
Subject: Max tables in a cluster?
MIME-Version: 1.0
Content-Type: multipart/alternative;
	boundary="----=_Part_953_543013687.1680678159145"
Importance: Medium
User-Agent: Zoho Mail
Archived-At: 
 <https://www.postgresql.org/message-id/18750398329.1002ca2ba316.8870907212977424118%40zohocorp.com>
Precedence: bulk

------=_Part_953_543013687.1680678159145
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit

Hi admins, 


Is there a limit to number of tables that a cluster can support? 

At the moment, one of my clusters has about 250k user tables (plus 150k toast tables) across a hundred schemas within a single database. 

My application team wants to double the number of schemas (and with it the tables)

I am looking for both pros and cons of doing this and what tuning can help this cluster perform acceptably. 


My questions are 

1. Is there a limit to number of tables a cluster can support effectively?

2. My guess is autovacuum would have to be tuned. Current gucs for this are 

max_autovacuum_workers = 6

maintenance_work_mem = 1GB

autovacuum_vacuum_scale_factor = 0.2

autovacuum_vacuum_cost_delay = 20ms


Are there any potential problems that I need to think about ?


My machine specs are as below.

RAM 256GB

32-core CPU

7TB SSD


thanks

karthik
------=_Part_953_543013687.1680678159145
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"><html><head>=
<meta content=3D"text/html;charset=3DUTF-8" http-equiv=3D"Content-Type"></h=
ead><body ><div style=3D"font-size: 10pt;"><div>Hi admins, <br></div><div><=
br></div><div>Is there a limit to number of tables that a cluster can suppo=
rt? <br></div><div>At the moment, one of my clusters has about 250k user ta=
bles (plus 150k toast tables) across a hundred schemas within a single data=
base. <br></div><div>My application team wants to double the number of sche=
mas (and with it the tables)<br></div><div>I am looking for both pros and c=
ons of doing this and what tuning can help this cluster perform acceptably.=
 <br></div><div><br></div><div>My questions are <br></div><div>1. Is there =
a limit to number of tables a cluster can support effectively?<br></div><di=
v>2. My guess is autovacuum would have to be tuned. Current gucs for this a=
re <br></div><div>max_autovacuum_workers =3D 6<br></div><div>maintenance_wo=
rk_mem =3D 1GB<br></div><div>autovacuum_vacuum_scale_factor =3D 0.2<br></di=
v><div>autovacuum_vacuum_cost_delay =3D 20ms<br></div><div><br></div><div>A=
re there any potential problems that I need to think about ?<br></div><div>=
<br></div><div>My machine specs are as below.<br></div><div>RAM 256GB<br></=
div><div>32-core CPU<br></div><div>7TB SSD<br></div><div><br></div><div>tha=
nks<br></div><div>karthik<br></div></div><br></body></html>
------=_Part_953_543013687.1680678159145--