Greetings All,
- I wanted to know how concurrency is implemented in GridFTP? - I wanted to know if you could point me to the exact code where concurrency is implemented and utilized? - I wanted to know when concurrent GridFTP control sessions/GridFTP server processes are started (based on the concurrency parameter specified in the globus-url-copy client), if each GridFTP Server Process have the same IP Address and Port # or do each of the concurrent GridFTP Server Process just have the same IP Address but different port number? For example, I performd a third party transfer file transfer using computers: WS12, WS11 and WS7, as seen below. So in this case will each of w12's control channels connect to w7 on port 5021? Or will one control channel connect to port 5021, the next one to 5022, the following one to port 5023, etc. ? *GridFTP servers were started on computers: WS12 and WS7* globus-gridftp-server -aa -p 5021 -log-level all -logfile /home/logfile.txt *Globus-url-copy was executed on computer WS11* globus-url-copy -vb -p 1 -cc 4 -dbg ftp://didclab-ws12.cse. buffalo.edu:5021/home/2GB_File.dat <ftp://didclab-ws12.cse.buffalo.edu:5021/home/lrodolph/2GB_File.dat> ftp://didclab-ws7.cse.buffalo.edu:5021/home/2GB_File_Copy.dat <ftp://didclab-ws7.cse.buffalo.edu:5021/home/lrodolph/2GB_File_Copy.dat> >> throughput.txt Note, I did read through many GridFTP documentations as well as traverse the code, but I have the above questions. this was a great overview https://docs.globus.org/faq/transfer-sharing/#what_are_gridftp_concurrency_pipelining_parallelism_and_striping Thank you in advance for your response, -- L. PhD Candidate
