Description
Related product team resources: N/A
Related product discussion: PRD link with explanations.
Motivation
Sometimes we see databases grow a lot but we don’t have the detailed information on what takes space. We also see long indexing times and the size of the database can explain a lot about this timings.
To alleviate this issue we simply display the size of the databases in the batches and the associated diff after a batch has been indexed..
Usage
Refer to the page of the experimental feature
// This is in the `/batches` - `results.stats.internalDatabaseSizes`.
"internalDatabaseSizes": {
"main": "1.36 MiB (+1.34 MiB)",
"external_documents_ids": "944 KiB (+944 KiB)",
"word_docids": "9.11 MiB (+9.11 MiB)",
"word_prefix_docids": "3.22 MiB (+3.22 MiB)",
"exact_word_prefix_docids": "16 KiB (+16 KiB)",
"word_pair_proximity_docids": "100.69 MiB (+100.69 MiB)",
"word_position_docids": "17.52 MiB (+17.52 MiB)",
"word_fid_docids": "10.31 MiB (+10.31 MiB)",
"word_prefix_position_docids": "9.91 MiB (+9.91 MiB)",
"word_prefix_fid_docids": "4.47 MiB (+4.47 MiB)",
"field_id_word_count_docids": "240 KiB (+240 KiB)",
"documents": "25.41 MiB (+25.41 MiB)"
}
Impacted teams
@meilisearch/docs-team, @meilisearch/integration-team
The new internalDatabaseSizes
field is in the batches stats.