On Tue, Jun 9, 2009 at 11:55 AM, Nicolas Pitre<nico@xxxxxxx> wrote: > > Do you clearly have a set of people for whom the clone works and another > set with issues? Yes. Several us are able to reliably clone every time. A few users are able to clone once or twice (maybe 1 out of every 10 times), but otherwise, it fails when the server drops the connection. >From what they said, it looked like read_in_full() (in wrapper.c) isn't getting a full buffer. In two cases, 4077 or 4082 out of 4097 bytes. xread is returning 0 as you'd expect for an EOF. > 147 KiB/s is not particularly slow either. What is fast enough for the > clone to usually succeed? Mine is 500 KB/sec where it always succeeds. I can try using dummynet to simulate a slow link though. > Also, since this is through an ssh connection, do you have anything in > your syslog from sshd (usually in /var/log/secure) about anything > unusual for the connection to go down? I'm trying to get that from the server admin now. -- To unsubscribe from this list: send the line "unsubscribe git" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html