Import and export Cassandra tables
Export tables to CSV or a typed dump, import CSV with column mapping and a preview, and restore a dump with TTL and WRITETIME.
Tablyne offers three ways to move table data, all from the table menu.
CSV export writes the whole table to a file, like cqlsh COPY TO. You choose the delimiter, the text used for NULL, and whether complex types are written as CQL literals or as JSON. An optional row limit stops the export early, and the result says when the limit was reached. Formula protection is on by default for CSV and can be turned off in the export dialog.
CSV import reads a file into the table, like cqlsh COPY FROM. You say whether the first row is a header, map each CSV column to a table column, and preview the result before importing. Every primary key column must be mapped. Cells with errors are shown in the preview, and you fix them first. You can set the consistency, a TTL in seconds, the rows per batch, and “only insert if it doesn’t exist”. If the import stops on an uncertain result, Tablyne tells you the row and asks you to check what was already written before you retry.
The typed dump is for a faithful copy. It writes the whole table to a JSON-lines file, with the type of each value and the TTL and WRITETIME of each cell. If the file already exists, Tablyne asks before replacing it. Non-frozen collections and UDTs do not carry a TTL or WRITETIME, and the app says so.
Restore reads such a dump back into the same table. Each row is inserted with INSERT ... IF NOT EXISTS, so an existing row is a conflict and is never overwritten. Tablyne keeps the remaining TTL and the original WRITETIME, skips rows that have already expired, and reports how many. It checks that the dump belongs to the table, and it does not restore counter tables, because adding again does not reproduce the value. Restore asks you to type the table name to confirm.
From the console, results export as CSV, JSON, JSONL, Markdown or XLSX. Copying a table to another connection is not available for Cassandra and ScyllaDB.