Dave> Try using "-W" to disable the rsync rolling checksum algorithm
Dave> when copying between two NFS mounts, because that causes extra
Dave> NFS traffic. Rsync's algorithm is optimized for minimizing
Dave> network traffic between its two halves at the expense of extra
Dave> local access and in your case the "network" is between processes
Dave> on the same machine and the "local" is over a network.
Here's some more info on the problem I've been having doing an rsync
between two NetApp toasters, running on one Solaris 2.6 server with a
direct connection to each toaster.
After chatting with Dave D. about this, I've added the following
options:
-W and --timeout=1800
to my script to see if that made any difference, and it didn't. I got
up this morning and found one rsync setup stuck in the same old
waitid() loop (2.4.6 this time btw).
But the interesting thing was that instead of the normal three
processes just sitting there, one in waitid() loop and the other two
in poll() loops, there was just two processes, one each in poll() and
waitid(), so I think that the problem is not in the timeouts.
If need be, I'll apply patches to my Solaris box to make this work
better. My NetApps are both running DataOnTap 5.3.6R2, which is
fairly recent, though not the latest version.
Any other hints?
Thanks,
John
John Stoffel - Senior Unix Systems Administrator - Lucent Technologies
[EMAIL PROTECTED] - http://www.lucent.com - 978-952-7548
