Re: Why could different data in a table be processed with different performance?

From: Vladimir Ryabtsev <greatvovan(at)gmail(dot)com>
To: pryzby(at)telsasoft(dot)com
Cc: f(dot)pardi(at)portavita(dot)eu, pgsql-performance(at)lists(dot)postgresql(dot)org
Subject: Re: Why could different data in a table be processed with different performance?
Date: 2018-09-25 01:11:12
Message-ID: CAMqTPqn=N--+MOr81-5qXvt3zpib-1TURmLjn27eHYj6b6L-qw@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

> did you try either 1) forcing a bitmap scan (of only one index), to force
the heap reads to be ordered, if not sequential? SET enable_indexscan=off
(and maybe SET enable_seqscan=off and others as needed).
Disabling index scan made it bitmap.
It is surprising, but this increased read speed in both ranges.
It came two times for "fast" range and 3 times faster for "slow" range (for
certain segments of data I checked on, the whole experiment takes a while
though).
But there is still a difference between the ranges, it became now ~20 MB/s
vs ~6 MB/s.

Vlad

In response to

Browse pgsql-performance by date

  From Date Subject
Next Message Justin Pryzby 2018-09-25 02:19:54 Re: Why could different data in a table be processed with different performance?
Previous Message Vladimir Ryabtsev 2018-09-25 00:59:12 Re: Why could different data in a table be processed with different performance?