In looking at my diffengine instances with large SQLite databases, by using py-spy I noticed that they were spending almost all their time executing the versions SQL query. Since we only use one result, adding limit(1) to the query may significantly improve performance (particularly for large databases).
In looking at my diffengine instances with large SQLite databases, by using py-spy I noticed that they were spending almost all their time executing the
versions
SQL query. Since we only use one result, addinglimit(1)
to the query may significantly improve performance (particularly for large databases).