Running the Refract executable, I have noticed that there is a memory leak on the order of 10MB per file processed.
When dealing with 5-minute data, this quickly causes the Refract program to crash with an error code if you attempt to process more than ~36 hours of data per run.
I have a workaround that partitions my jobs into batches that are a maximum of 24 hours long. Presently, there appears to be a memory leak, at least in the CentOS 7 .rpm.
Steps to reproduce:
I have a workaround that partitions my jobs into batches that are a maximum of 24 hours long. Presently, there appears to be a memory leak, at least in the CentOS 7 .rpm.