I have a large table with about 50 billion records. The files to be loaded are not in CSV format. The files are processed in java and the data uploaded to the database using COPY INTO through the MapiSocket interface. I do not have the record count readily available. My question is this: If I open multiple MapiSockets and upload multiple streams of data at the same time to the same table, is this supported and stable? I am asking because the database appears to become unresponsive after the bulk load. (mclient seems to hang on connect) If I stop and the start the database after the bulk load, I get: "failed to fork mserver: database 'test_db' has inconsistent state (sabaoth administration reports running, but process seems gone)" -- View this message in context: http://old.nabble.com/Parallel-COPY-INTO-tp32503834p32503834.html Sent from the monetdb-users mailing list archive at Nabble.com.