Hi, I am trying to copy documents between databases using MLCP. The source ML host is ML7 and destination ML host is ML9 and are installed in AWS. I am running the MLCP script from an AWS EC2 instance.
As per the below report for MLCP, it looks like the copy between databases is slower when compared to the XQSync process. Because XQSync took only half of the MLCP time. INFO contentpump.LocalJobRunner: com.marklogic.mapreduce.MarkLogicCounter: INFO contentpump.LocalJobRunner: ESTIMATED_INPUT_RECORDS: 4459979 INFO contentpump.LocalJobRunner: INPUT_RECORDS: 3589532 INFO contentpump.LocalJobRunner: OUTPUT_RECORDS: 3589532 INFO contentpump.LocalJobRunner: OUTPUT_RECORDS_COMMITTED: 3587928 INFO contentpump.LocalJobRunner: OUTPUT_RECORDS_FAILED: 1604 INFO contentpump.LocalJobRunner: Total execution time: 58989 sec I read the MLCP documentation and tried few changes like thread_count, batch_size, -fastload etc., But still could not copy in faster rate. Also, the input, output records and its estimates is not looking correct and same. Can someone please help me to figure out what can be done to make it faster. MLCP Script: copy -mode local -input_host aws-host1 -input_port 12345 -input_username username -input_password password -input_database TEST_DB -output_host aws-host2 -output_port 12345 -output_username username -output_password password -output_database TEST_DB -fastload true -collection_filter TEST -batch_size 100 Thanks, Prasanth
_______________________________________________ General mailing list General@developer.marklogic.com Manage your subscription at: http://developer.marklogic.com/mailman/listinfo/general
