[OpenAFS-devel] Re: fsync/nofsync performance numbers on Solaris

[Dale Ghent]

On Fri, September 15, 2006 12:39 pm, Robert Banz wrote:

> I didn't bother testing zfs with the fsync() on, I knew it would be
> pretty horrible in light of the current fsync "bug" it's got...

Yeah, this bug should be seeing a patch soon enough. It has been fixed in
Nevada since early July IIRC. This fsync() bug in ZFS is a doozie...
whenever a fsync() is issued to a file handle, the ZFS bug causes all
other opened files to be fsync'd as well. This kills NFS performance and
obviously would nail AFS namei, too.

/dale

-- 
Dale Ghent
UNIX Systems Specialist
UMBC - Office of Information Technology
ECS 201 - x51705