The mbuffer on the recieving size does fill up, but at that point it's sustaining 270MB/sec through put which I'm happy with.
What I'm unhappy about is that mbuffer bombs before the proccess is completed with a broken pipe error shown on both sides. Several folks have indicated that they have not experianced these issues with mbuffer, so I was wondering if it might have something to do with my 10Gbit cards. Both sides are 10x256GB SSDs in a raidz2 vdev. testing with bonnie++ showed that the pools can write at over 450MB sec. I think the bottleneck when using mbuffer may be in the pipe process used to moved data from mbuffer to zfs receive. Richard J. -- This message posted from opensolaris.org _______________________________________________ networking-discuss mailing list networking-discuss@opensolaris.org
