Search Postgresql Archives

Re: Accounting for between table correlation

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 1/15/21 9:19 AM, Alexander Stoddard wrote:
I am having ongoing trouble with a pair of tables, the design of which is beyond my control.

There is a 'primary' table with hundreds of millions of rows. There is then a 'subclass' table ~ 10% of the primary which has additional fields. The tables logically share a primary key field (although that is not annotated in the schema).

Membership of the subclass table has high correlation with fields in the primary table - it is very much not random. It seems query plans where the two tables are joined are 'unstable'. Over time very different plans can result  even for unchanged queries and some plans are exceedingly inefficient.

I think what is going on is that the query planner assumes many fewer rows are going to join to the subtable than actually do (because of the strong correlation).

Can anyone offer any advice on dealing with this scenario (or better diagnosing it)?

Do the tables get analyzed on a regular basis?

--
Angular momentum makes the world go 'round.





[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Index of Archives]     [Postgresql Jobs]     [Postgresql Admin]     [Postgresql Performance]     [Linux Clusters]     [PHP Home]     [PHP on Windows]     [Kernel Newbies]     [PHP Classes]     [PHP Books]     [PHP Databases]     [Postgresql & PHP]     [Yosemite]

  Powered by Linux