[Gluster-devel] AFR between two bricks over 3000 miles

Anand Avati avati at zresearch.com
Thu Feb 28 05:21:30 UTC 2008

using a combination of write-behind with io-threads (with option
flush-behind on) prevents the wait of upto N-MB of data (where N is 'option
cache-size NMB' of io-threads)


2008/2/28, nathan at robotics.net <nathan at robotics.net>:
> Is it possible to afr between two bricks in CA and NY with decent
> response time with each system? The hosts are configured as client and
> server, but so far my tests have been with afr on the client side only.
> In testing I am getting very slow write times, my hope was with AFR was
> writes / reeds would be close to the speed of local since I am using the
> NUFA scheduler in the client.
> Should I move AFR and NUFA to the server rather then the client? Ideally,
> writes would hit the local box at its write speed and start transfer data
> to the remove AFR until finished no slowing down local response time. Now
> I understand that may not be possible, but what about at least having
> say an 8 meg window so the system would be much more responsive with
> smaller files, yet not get stuck with two sites modifying the same file.
> Thanks for the help, there has got to be a better solution then
> YottaYotta.
> ><>
> Nathan Stratton
> nathan at robotics.net
> http://www.robotics.net
> _______________________________________________
> Gluster-devel mailing list
> Gluster-devel at nongnu.org
> http://lists.nongnu.org/mailman/listinfo/gluster-devel

If I traveled to the end of the rainbow
As Dame Fortune did intend,
Murphy would be there to tell me
The pot's at the other end.

More information about the Gluster-devel mailing list