[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]
Re: porting ZFS
Hello. While I'm not an expert with ZFS, I can think of several
reasons why it would be useful to have for NetBSD.
1. With NetBSD acting as an ISCSI target, giving NetBSD the ability to
provide large amounts of storage to serve via that mechanism would be very
2. In Xen environments, it would be useful to be able to address large
amounts of robust storage. Again, ZFS seems like a good way to do this.
I'm talking for dom0, in case it isn't clear.
LFS is nice, but I don't see LFS and ZFS as addressing the same issues
at all. LFS might be useful as a filesystem stored on top of a xen stack,
so that the domU can put files on an ISCSI block device, but ZFS would
allow the dom0 to address a lot of disk so that it can provide those large
block devices to the domU's.
Hope that helps.
On Sep 17, 9:14pm, "Neel Sheyal" wrote:
} Subject: Re: porting ZFS
} > Personally, I'm not really sure if ZFS is appropriate as a general-purpose
} > system for NetBSD, as it is a complex and very hardware-demanding beast. And
} > porting seems to be hard - at least Matthew Dillon found it easier and
} > to code a new state-of-the-art (and well-portable!!) BSD file system
} > porting and improving ZFS:
} > http://apollo.backplane.com/DFlyMisc/hammer01.pdf
} I must admit that I am not a pundit of ZFS. So, I cannot compare
} intelligently ZFS with Hammer. But if someone as experienced like
} Matthew labeled it "hard", then it might be worth revisiting the
} effort. There might be valid reasons for it.
} It would be great if all the masters chimed in with their views as to
} why NetBSD might benefit having ZFS. And this is very important before
} we dedicate cycles to porting ZFS.
} > As a matter of principle, one goal shall be to complete unfinished
} Mark: while the debate rages, do you have specific FS related project
} that need immediate attention.
>-- End of excerpt from "Neel Sheyal"
Main Index |
Thread Index |