Skip to content

Instantly share code, notes, and snippets.

@v-kolesnikov
Created December 26, 2022 09:26
Show Gist options
  • Save v-kolesnikov/316c1911f2ea2fc25cd012721d56559c to your computer and use it in GitHub Desktop.
Save v-kolesnikov/316c1911f2ea2fc25cd012721d56559c to your computer and use it in GitHub Desktop.
opts[:max_dataset_size]
It is a very important constant that provides a useful performance optimization.
For example we have N records in `users` table and want to show that number
under the corresponding table on `/users` page. It would be quite fast for DB
to count some tausands records and return that value back to the application.
But in case of 1M records the same task will load the DB server significantly
and without the relevant reason. There is no difference for web-app purposes
whether we have 1M or 1M+1 records in the DB .
And here `max_dataset_size` helps us.
Instead of counting the entire table as:
ds = User.dataset
ds.count
SELECT count(*) from users;
=> Counting 1M records ...
we count items in a limited dataset:
ds = User.dataset.limit(10000)
ds.count
SELECT count(*) FROM (SELECT * from users LIMIT 10000)
=> <= 10000
Then we can compare returned value with ds.pagination_record_count
that always be <=10000 since ds is limited with that value.
If count is lower than max_dataset_size - we know the exactly count of
records. Otherwise we are sure there are more than 10000 items.
Finally this option serve for fast UI and better user experience.
In real life there is hardly a case where web ui should operate with
large dataset with 10k+ records.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment