On Tue, 6 May 2008, Christopher Hawkins wrote:
Thanks, I was wondering about that. Below I suggested using glusterfs to do the copy and forgetting about an external rsync, but I didn't know how that would perform compared to an rsync. Looks like your data confirms that if the xattrs get set correctly, allowing gluster to do the copying / creating would be a good fast solution.
Indeed, to achieve the copy as fast as possible, I copied all the files to glusterfs with the 2nd server disconnected (it was effectively local disk-to-disk copy), then when that was up and running I just brought the 2nd server up and did a recursive find sync. I did it in parallel for all first level subdirectories (~ 5) to improve bandwidth saturation due to latencies on smaller files. I was quite pleasantly surprised by how low the overheads seem to be.
Gordan
