Re: FUSE Call for Testing

On Fri, Aug 9, 2019 at 3:02 AM Gary Jennejohn <gljennjohn_at_gmail.com> wrote:
>
> On Thu, 8 Aug 2019 12:34:52 -0600
> Alan Somers <asomers_at_freebsd.org> wrote:
>
> [snip]
>
> > VM images: http://ftp0.nyi.freebsd.org/pub/FreeBSD/snapshots/VM-IMAGES/13.0-CURRENT/amd64/20190808/
> > ISOs: http://ftp0.nyi.freebsd.org/pub/FreeBSD/snapshots/ISO-IMAGES/13.0/
> >
> > Thanks for any feedback you can give!
>
> So, I tried it with fuse-ext2 (mount RO) and ntfs-3g
> (fusefs-ntfs).
>
> Both work.  Interrupting a transfer also doesn't seem to cause
> problems.
>
> Reading from NTS was pretty fast at >= 65MiB/sec.  I suspect that
> most FBSD users will use NTFS in this way.
>
> Hard to say whether writing to NTFS is faster than it was before
> because the transfer rate bounces around all over the place.
> Sometimes a few 100KiB, sometimes 10s of MiB.  I suspect that the
> small transfers are associated with filesystem bookkeeping.
> Small tranfers are most frequent.
>
> NOTE that I had to reinstall fusefs-ntfs (I had a really old
> version) for it to work.  I tried it with and without LOCK and
> UBLIO but no real difference in speed was noticeable.
>
> NOTE2 that writing to NTFS was tested by copying a 5.2GiB file
> from a SATA SSD to a 64GiB USB3 stick plugged into a USB3 port.
> I then used /usr/bin/sum to test integrity.  The result was
> identical for both files, so that's good news.
>
> --
> Gary Jennejohn (gj_at_)

Thanks for the report, Gary.  BTW, fusefs mounts are only
interruptible when mounted with "-o intr".  If you didn't use that
option, then the signal would only interrupt cp after the write to
fusefs was done.  Also, not every fuse file system supports
interrupts.  Looking through its sources, I don't think that
fusefs-ntfs does.
-Alan