Urgent! When sed processes ultra-large text files (21 GB), the memory is insufficient ..

Urgent! When sed processes ultra-large text files (21 GB), the memory is insufficient .. -- Linux general technology-Linux technology and application information. For details, refer to the following section. I want to extract six specific columns from a table in the database. The current size is 21 GB. Because some of the exported text files also have line breaks in each record, I used the composite command ";" to achieve this. Then extract the target column with the MPs queue and awk. However, after the command is knocked out, the memory consumption soared and the system had to stop. I want to ask if the sed pipeline operation has been damaged when I removed the line feed (more than 80 million records are still removed after the unexpected line feed) in the first step. Is there any good suggestions.