Home/ Questions/Q 8092711
In Process

The Archive Base Latest Questions

Editorial Team
  • 0
Asked: 2026-06-05T20:20:43+00:00

I’ve got a table like so: CREATE TABLE IF NOT EXISTS grades(_id, timestamp, extra);

  • 0

I’ve got a table like so:

CREATE TABLE IF NOT EXISTS grades(_id, timestamp, extra);

I want to create an index on “timestamp”, so I’m doing:

CREATE INDEX idx_timestamp ON grades(timestamp);

I want to select 20 records at a time based off the timestamp then:

SELECT * FROM grades WHERE timestamp > 123 ORDER BY timestamp ASC LIMIT 20;

So, is there a more efficient way I can define the column “timestamp”? I’m just guessing that specifying it as an indexed column is all we can do, and specifying “ASC” for sort order is a no-op – or can I tell sqlite to store records presorted by timestamp in the first place?

I’m basically trying to implement a paging system, selecting a chronologically ordered page of 20 items at a time.

Thanks

Share

Leave an answer

You must login to add an answer.

Need An Account,

1 Answer

  1. Editorial Team

    This is fine. It will use the index to order, so it will increase the speed. although, depending what you are doing you will may want to cache some records.

    SELECT * FROM grades WHERE timestamp > 123 ORDER BY timestamp ASC LIMIT 200;

    We can now get 200 at a time, but on we now will let javascript or the server handle the paging. Basically you would keep track of where the paging is and then only when needed hit the database again. Also, by using the more limited WHERE clause for the timestamp, it’s actually quite fast and efficient. Better than using the LIMIT N,M.

    If you do the second option, you will be able to cache that query as well, depending how often it gets hit. So, if multiple people keeping querying that some thing, the database will cache it and it will come back really fast since it’s already there.

    • 0