Jake Van Meter

It looks like the reason things are so slow is that every record is getting a conflict, which takes additional time to resolve. Do you have something like a datetime or timestamp in the primary key, as that could cause something like this. Can you post the table defenition here?

You can turn off the estimated row count (initial.load.use.estimated.counts) to get a more accurate reading of how many rows are there. You could also run Analyze table table_name replacing table_name with the name of your table to reset the statistics.

Andrew, Snowflake is a Pro only database, as seen on this page in the SymmetricDS website: https://symmetricds.org/connectors/ In theory it could be possible, but you would have to setup Snowflake as load only and it would be incredibly slow. The Pro version has optimized this to be the fastest possible implementation.

Below is an article that I think would help answer your question: https://www.jumpmind.com/blog/blog/data-trends/common-data-replication-mistakes/ Specificially, take a look at the Conflicted Data section and the solutions it suggests.

Below is a link to aid with understanding how transforms work and the best way to set them up. https://www.jumpmind.com/blog/blog/how-to/symmetricds-transform-pipeline/ Hopefully this answers your question!

Can you try putting COST as the target_column_name in your insert into SYM_TRANSFORM_COLUMN. It would look like this: insert into SYM_TRANSFORM_COLUMN ( transform_id, include_on, target_column_name, source_column_name, pk, transform_type, transform_expression, transform_order, last_update_time, last_update_by, create_time ) values ( 'itemSellingPriceTransform','*','COST','COST', 0, 'remove', '', 1, current_timestamp, 'Documentation', current_timestamp );

Currently that is not something that is being explored. If you are looking to have this functionality added, you could submit a pull request with the proposed changes. Alternatively, you could create the SQLite database ahead of time with the user already setup and supply that user to SymmetricDS.

The database URL that you give SymmetricDS for PostgreSQL is the database that SymmetricDS will use. PostgreSQL does not allow cross database connections. You will need to setup a separate node if you are attempting to use a database that is different than the one in the database URL you provided initially.