My top queries to analyze the DB on PostgreSQL

Hi community, 


Today I would like to share a small snippet to investigate DB and RDBMS PostgreSQL. 

Before starting to check and do “explain analyze” of slowest and frequent slow queries I do:

select sum(xact_commit) as commits, sum(xact_rollback) as rollbacks
from pg_stat_database;

Just to understand how often we meet with rollbacks.


I do recommend to run a few times that query with a small timeout to understand frequency. 


Next one is used and remained connections query:

select max_conn,used,res_for_super,max_conn-used-res_for_super res_for_normal
 (select count(*) used from pg_stat_activity) t1,
 (select setting::int res_for_super from pg_settings where name=$$superuser_reserved_connections$$) t2,
 (select setting::int max_conn from pg_settings where name=$$max_connections$$) t3;


After we are ready to check the status of connections. Helpful to plan pooling.

with states as
   (select datname, client_addr, case
       when now() - state_change < interval '10 seconds' then '10sec'
       when now() - state_change < interval '30 seconds' then '30sec'
       when now() - state_change < interval '60 seconds' then '60sec'
       else 'idle' end
   as stat from pg_stat_activity)
select datname, client_addr, stat, count(*)
from states group by datname, client_addr, stat;

It helps to understand the counts and duration. 


Then I do a review of quantitative parameters of databases usually of Atlassian systems. 

It’s size of database, number of rows and indexes. 

-- This query returns list of largest (by data size) tables.
select schemaname as table_schema,
relname as table_name,
pg_size_pretty(pg_total_relation_size(relid)) as total_size,
pg_size_pretty(pg_relation_size(relid)) as data_size,
pg_size_pretty(pg_total_relation_size(relid) - pg_relation_size(relid))
as external_size
from pg_catalog.pg_statio_user_tables
order by pg_total_relation_size(relid) desc,
pg_relation_size(relid) desc
limit 10;


All those queries you can checkout from repo.

Hope it helps. 

Feel free to share some interesting snippets of DB, RDBMS analysis.


Popular posts from this blog

Overview and practical use cases with open source tracing tool for Java Apps. Glowroot. Installation

How only 2 parameters of PostgreSQL reduced anomaly of Jira Data Center nodes

Atlassian Community, let's collaborate and provide stats to vendors about our SQL index usage