Databases
StackOverflow2013
Postsdbo
POSpeeding up SQL query when sorting on foreign keys
 

Note that there are some explanatory texts on larger screens.

plurals
  1. POSpeeding up SQL query when sorting on foreign keys
    primarykey
    data
    text
    <p>This is more of a generic SQL problem but I'm using Firebird 2.5 if anyone knows of a Firebird/Interbase specific optimization. First, the following is a simplified example schema to illustrate the issue I'm trying to solve:</p> <pre><code>CREATE TABLE users ( id INTEGER PRIMARY KEY, name VARCHAR(16) ); CREATE TABLE data_set ( id INTEGER PRIMARY KEY, name VARCHAR(64) ); CREATE UNIQUE INDEX data_set_name_idx ON data_set(name); CREATE TABLE data ( user_id INTEGER, data_set_id INTEGER, data BLOB, PRIMARY KEY(user_id, data_set_id) ); CREATE INDEX data_user_id_idx ON data(user_id); CREATE INDEX data_data_set_id_idx ON data(data_set_id); </code></pre> <p>The query I'm trying to run is as follows:</p> <pre><code>SELECT users.name, data_set.name, data FROM users, data_set, data WHERE user_id=XXX AND user_id=users.id AND data_set_id=data_set.id ORDER BY data_set.name; </code></pre> <p>With 'XXX' being filled in with the <em>user_id</em> I want. So what I'm doing is selecting all the rows from the <em>data</em> table that are owned by a particular user and I'm sorting the results based on the <em>data_set</em> name.</p> <p>This works as it is but the problem is the <em>data</em> table has over a billion rows in it and the <em>data_set</em> table is not small either. The result set for a single user id may be many hundreds of millions of rows. What happens is that in order for the ORDER BY to work the database has to create a massive amount of temporary data which is incredibly slow and uses a lot of disk space. Without the ORDER BY it's fast but obviously not sorted like I need.</p> <p>One solution would be to take the <em>data_set.name</em> values and just put them in a varchar column in <em>data</em>. Then that could be indexed and would be quick to sort. The problem with this approach is that it will have a lot of duplicate data and make the database absolutely massive.</p> <p>Another solution would be something like an Indexed View or an indexed Computed Column. As far as I know neither of those is supported by Firebird.</p> <p>Any other ideas?</p>
    singulars
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. PTQuestion
    1. USHugues Van Landeghem
    1. USCR.
    plurals
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. PO
      singulars
      1. PTAnswer
    2. PO
      singulars
      1. PTAnswer
    3. PO
      singulars
      1. PTAnswer
    1. VO
      singulars
      1. POSpeeding up SQL query when sorting on foreign keys
      1. This table or related slice is empty.
      1. VTUpMod
    2. VO
      singulars
      1. POSpeeding up SQL query when sorting on foreign keys
      1. USDenis Otkidach
      1. VTFavorite
    3. VO
      singulars
      1. POSpeeding up SQL query when sorting on foreign keys
      1. This table or related slice is empty.
      1. VTUpMod
    1. COI don't have a definite answer, but I think you can play around more with your indexes and benchmark them with the query.
      singulars
      1. POSpeeding up SQL query when sorting on foreign keys
      1. USo.k.w
    2. COWhy do you want to pull "hundreds of millions of rows" for a single user in the first place? It will take time not only to process the query but to transfer the data to the client.
      singulars
      1. POSpeeding up SQL query when sorting on foreign keys
      1. USliggett78
    3. CO@liggett78: Oh, I agree it will take time to transfer it because it's a lot of data. That's why I'm trying to speed this up as much as possible. It doesn't need to be as slow as it is because it's wasting a lot of time creating temporary data for the sort.
      singulars
      1. POSpeeding up SQL query when sorting on foreign keys
      1. USCR.
 

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