SQLite EF Core Database Provider Limitations
The SQLite provider has a number of migrations limitations. Most of these limitations are a result of limitations in the underlying SQLite database engine and are not specific to EF.
The common relational library (shared by Entity Framework relational database providers) defines APIs for modelling concepts that are common to most relational database engines. A couple of these concepts are not supported by the SQLite provider.
- Computed columns
SQLite doesn't natively support the following data types. EF Core can read and write values of these types, and querying for equality (
where e.Property == value) is also supported. Other operations, however, like comparison and ordering will require evaluation on the client.
DateTimeOffset, we recommend using DateTime values. When handling multiple time zones, we recommend converting the values to UTC before saving and then converting back to the appropriate time zone.
Decimal type provides a high level of precision. If you don't need that level of precision, however, we recommend using double instead. You can use a value converter to continue using decimal in your classes.
modelBuilder.Entity<MyEntity>() .Property(e => e.DecimalProperty) .HasConversion<double>();
The SQLite database engine does not support a number of schema operations that are supported by the majority of other relational databases. If you attempt to apply one of the unsupported operations to a SQLite database then a
NotSupportedException will be thrown.
Migrations limitations workaround
You can workaround some of these limitations by manually writing code in your migrations to perform a table rebuild. A table rebuild involves renaming the existing table, creating a new table, copying data to the new table, and dropping the old table. You will need to use the
Sql(string) method to perform some of these steps.
See Making Other Kinds Of Table Schema Changes in the SQLite documentation for more details.
In the future, EF may support some of these operations by using the table rebuild approach under the covers. You can track this feature on our GitHub project.