Databases
StackOverflow2013
Postsdbo
PO
 

Note that there are some explanatory texts on larger screens.

plurals
  1. PO
    primarykey
    data
    text
    <p>If the query planner makes bad decisions it's mostly one of two things:</p> <h2>1. The <strong>statistics</strong> are off.</h2> <p><sub>Meaning "inaccurate", not "turned off".</sub></p> <p>Do you run <code>ANALYZE</code> enough? Also popular in it's combined form <code>VACUUM ANALYZE</code>. If <a href="http://www.postgresql.org/docs/current/static/routine-vacuuming.html#AUTOVACUUM" rel="noreferrer">autovacuum</a> is on (which is the default in modern-day Postgres), <code>ANALYZE</code> is run automatically. But consider:</p> <ul> <li><a href="https://dba.stackexchange.com/questions/18664/are-regular-vacuum-analyze-still-recommended-under-9-1/18694#18694">Are regular VACUUM ANALYZE still recommended under 9.1?</a></li> </ul> <p><sup>(Top two answers still apply for Postgres 9.6.)</sup></p> <p>If your table is <em>big</em> and data distribution is <em>irregular</em>, raising the <a href="http://www.postgresql.org/docs/current/interactive/runtime-config-query.html#RUNTIME-CONFIG-QUERY-OTHER" rel="noreferrer"><code>default_statistics_target</code></a> may help. Or rather, just <a href="http://www.postgresql.org/docs/current/interactive/sql-altertable.html" rel="noreferrer">set the statistics target</a> for relevant columns (those in <code>WHERE</code> or <code>JOIN</code> clauses of your queries, basically): </p> <pre><code>ALTER TABLE ... ALTER COLUMN ... SET STATISTICS 1234; -- calibrate number </code></pre> <blockquote> <p>The target can be set in the range 0 to 10000;</p> </blockquote> <p>Run <code>ANALYZE</code> again after that (on relevant tables). </p> <h2>2. The <strong>cost settings</strong> for planner estimates are off.</h2> <p>Read the chapter <a href="http://www.postgresql.org/docs/current/static/runtime-config-query.html#RUNTIME-CONFIG-QUERY-CONSTANTS" rel="noreferrer">Planner Cost Constants</a> in the manual. </p> <p>Look at the chapters <strong>default_statistics_target</strong> and <strong>random_page_cost</strong> on this <a href="http://wiki.postgresql.org/wiki/Tuning_Your_PostgreSQL_Server" rel="noreferrer">generally helpful PostgreSQL Wiki page</a>.</p> <p>Of course, there can be many other possible reasons, but these are the most common ones by far.</p>
    singulars
    1. This table or related slice is empty.
    1. POKeep PostgreSQL from sometimes choosing a bad query plan
      singulars
      1. PTQuestion
    1. PTAnswer
    1. USErwin Brandstetter
    1. USErwin Brandstetter
    plurals
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. POKeep PostgreSQL from sometimes choosing a bad query plan
      singulars
      1. PTQuestion
    1. This table or related slice is empty.
    1. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTUpMod
    2. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTAcceptedByOriginator
    3. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTUpMod
    1. COEven worse: the statistics/histograms for the {x,y,z} dimensions might be *not really independent*. Larger histograms *may* help, though.
      singulars
      1. PO
      1. USwildplasser
    2. COThanks for the suggestions. I have run `VACUUM ANALYZE` recently, but I tried again, and also after changing `default_statistics_target` to its maximum of 10000. Unfortunately, the wrong query plan was still chosen after doing that. Changing `random_page_cost` didn't have an effect either. Thanks for the suggested reading, I'll go through those pages carefully.
      singulars
      1. PO
      1. USMark Longair
    3. CO@wildplasser: Indeed, the x, y and z values certainly aren't independent. That makes me realise, though, that that the only cases I've seen of this misplanning would be at the very low end of the x values histogram...
      singulars
      1. PO
      1. USMark Longair
 

Querying!

 
Guidance
A row detail

Detail views are divided into sections. All the information in the data section comes from columns in the selected row. The other sections display data from other, related rows.

Related data can be related in a to-one or a to-many fashion. Captions of data related in a to-many fashion link to a list view showing a filtered view of the table.

Try moving around until you find a non-empty to-many entry and click on the label to get to one. You can move back to the root by clicking on the database name in the header.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload