For 2 million records, use a bulk/batch migration approach rather than creating records one at a time.
- Use bulk operations such as CreateMultiple and UpdateMultiple where supported to reduce API calls and improve throughput.
- Batch the data into manageable chunks and tune the batch size to avoid request limits and timeouts.
- Temporarily disable non-essential synchronous plug-ins and workflows during the migration to avoid unnecessary processing for every record.
- Handle duplicates deliberately. Bulk operations suppress duplicate detection by default, so clean the data beforehand or explicitly enable detection where appropriate.
- Run during off-peak hours and monitor API limits, throttling, failures, and overall environment performance.
- Plan relationships carefully, importing parent records before dependent child records when required.
Example: Split the 2 million records into manageable batches, validate each batch, process it using bulk operations, and retry failed batches rather than restarting the entire migration.
Interview tip: At this scale, the key is bulk operations + controlled batching + minimizing unnecessary automation, rather than simply increasing the import size.
