Explain rsync's --blocking-io option.

Question

I have problems with understanding the --blocking-io option in rsync. Here's the descripton from the man page:

"This tells rsync to use blocking I/O when launching a remote shell transport. If the remote shell is either rsh or remsh, rsync defaults to using blocking I/O, otherwise it defaults to using non-blocking I/O. (Note that ssh prefers non-blocking I/O.)"

• When using rsync as the transport, i.e. you don't use -e, does this have an effect at all?

• I guess the result of running rsync with or without that option is the same and the difference is in 'how' rsync does the transfer?

• What's the main reason I would wanted to use this option? Performance? Looks like rsync provides sensible defaults for me, so why would I want to use it at all?

• And the last one: What does "Note that ssh prefers non-blocking I/O" mean?

Answer 1

Sorry part of my research got mixed up. That is my fault. (Corrected or added stuff is in italics)

Q1: Yes. Rsync uses asynchronous I/O (non-blocking). By using that option, you are forcing it to use blocking I/O mode. This would be used when the remote shell (rsh) cannot handle non-blocking mode.

Q2: Yes. The end result will be the same.

Q3: You would want to use this option if the remote shell is not capable of handling non-blocking i/o

Q4: "Non-blocking mode makes sure when your connection dies, your screen session won't block ('hang')." (http://webcache.googleusercontent.com/search?q=cache:xmCnC0y5QU0J:www.warp13.co.uk/irssissh+ssh+blocking+mode&cd=13&hl=en&ct=clnk&gl=us&client=firefox-a) SSH just prefers that

I hope that helps!