<div class="gmail_quote">On Fri, Jan 8, 2010 at 2:45 PM, Carlos Carvalho <span dir="ltr"><<a href="mailto:carlos@fisica.ufpr.br">carlos@fisica.ufpr.br</a>></span> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
It seems the latest timeout changes in 3.1 work well.<br></blockquote></div><br>Great to hear! It's nice to know that a 3.1.0 receiver is indeed helping to keep things alive that used to die, even without 3.1.0 on the sending side (which is exactly what I was hoping for). A 3.1.0 sender also helps to keep a non-3.1.0 receiver going, as some testing I'm doing at work has proven (0 timeouts when doing some --dry-run difference-finding runs over about 7 TB of files to each of 30 mirrors -- that's a huge improvement). I'm liking the new 3.1.0 I/O and timeout code a lot.<br clear="all">
<br>..wayne..<br>