CYBERTEC PostgreSQL Logo

PostgreSQL: ltree vs. WITH RECURSIVE

06.2020 / Category: / Tags: |

After my last post about ltree and recursive data in PostgreSQL people have asked me privately about performance issues. To share this information, I decided to come up with a follow up post to discuss this topic in a bit more detail. WITH RECURSIVE in PostgreSQL is efficient. However, ltree does have its strengths as well. Let us take a closer look …

Preparing sample data

In preparation for a test I have created a table which contains a balanced tree:

There is a bit of magic involved here but let us not worry too much about the details at this point. Once the data is generated, you should see the following content in your table:

12 million rows will be a nice way to show how efficient ltree can really be. As stated before, using a recursive query on this data might take a bit more time than running the query on prepared data. What you can do in the next listing is to resolve the tree and materialize an ltree column containing all the data. Make sure ltree is enabled in your database:

Materializing tree data in PostgreSQL

Here is a materialized view containing the same data in a different format. The goal is to materialize the tree column to speed up searching:

To access the materialized view efficiently, we have to create a Gist index on the ltree column:

Note that a btree is not what we are looking for here. A gist index is necessary. The reason is that a btree only supports sorting which is not quite what we need here in this case.

Observing superior performance gains

If your data is static materializing makes a lot of sense. In case your data changes frequently and in case you always need 100% correct results it is of course pretty counterproductive to pre-calculate data. However, let us take a look at the speed of the materialized case:

The query executes in way less than a millisecond, which is exceptional.

Finally …

Performance is king and we have written a lot about PostgreSQL performance in the past. One of the key aspects is VACUUM. Read on to find out how to enable and disable AUTOVACUUM.


In order to receive regular updates on important changes in PostgreSQL, subscribe to our newsletter, or follow us on Facebook or LinkedIn.

One response to “PostgreSQL: ltree vs. WITH RECURSIVE”

  1. I'm afraid the GIST index has a limit of 255 items... Is there any method to use this functionality with very long data structures?

Leave a Reply

Your email address will not be published. Required fields are marked *

CYBERTEC Logo white
CYBERTEC PostgreSQL International GmbH
Römerstraße 19
2752 Wöllersdorf
Austria

+43 (0) 2622 93022-0
office@cybertec.at

Get the newest PostgreSQL Info & Tools


    This site is protected by reCAPTCHA and the Google Privacy Policy & Terms of Service apply.

    ©
    2024
    CYBERTEC PostgreSQL International GmbH
    phone-handsetmagnifiercrosscross-circle
    linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram