<div dir="ltr"><div class="gmail_extra"><div class="gmail_quote">On Mon, Oct 21, 2013 at 9:04 PM, Raymond Wagner <span dir="ltr"><<a href="mailto:raymond@wagnerrp.com" target="_blank">raymond@wagnerrp.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div class="im">On 10/21/2013 11:25 PM, Karl Newman wrote:<br>
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
so I ended up going back to mdadm with a simple BTRFS file system on<br>
top.<br>
</blockquote>
<br></div>
One of the core features of BTRFS is pervasive checksumming of data.<br>
Combined with internal redundancy, you have automatic scrubbing and<br>
repair of damaged files. That doesn't work if you externally handle<br>
redundancy through MDADM or hardware RAID.<br></blockquote><div><br>Interesting. I wanted it to work on the bare partitions, but it just wasn't mature. I was really looking at the high level features--ability to grow (and shrink), fast delete, efficient handling of large files, subvolumes, etc. Possibly ZFS may have been a better choice, but its license conflict with GPL means it will never be part of the Linux kernel, and I wanted something that I could use with a normal Linux boot CD (most of which aren't going to have ZFS support either as a module or FUSE).<br>
<br></div><div>Karl<br></div></div></div></div>