Re: Interesting problem with write data.

[Chris Mason <chris.mason@oracle.com>]

Excerpts from Tomasz Chmielewski's message of 2010-11-18 10:39:05 -0500:
> On 18.11.2010 16:07, Chris Mason wrote:
> 
> (...)
> 
> >> [27821.906513] btrfs-cache-8 D ffff88050c5fde98     0  8089      2 0x00000000
> >> [27821.906517]  ffff88051c3a9b60 0000000000000046 ffff88051c3a9b00 ffff88051c3a9fd8
> >> [27821.906522]  00000000000139c0 00000000000139c0 ffff88051c3a9fd8 ffff88051c3a9fd8
> >> [27821.906526]  00000000000139c0 ffff88050c5fde98 ffff88050c5fdea0 ffff88050c5fdb00
> >> [27821.906530] Call Trace:
> >> [27821.906534]  [<ffffffff8159fc4e>] io_schedule+0x5e/0xa0
> >> [27821.906538]  [<ffffffff81109f15>] sync_page+0x45/0x60
> >
> > So, you're caching block groups.  What you want to do is use Josef's new
> > block group caching code.
> >
> > mount -o space_cache /dev/xxx
> >
> > Do the test and let the caching threads finish, then unmount and then
> > your next run should be fast.
> 
> # mount -o space_cache /dev/sdb4 /mnt/btrfs/
> [29720.305741] btrfs: enabling disk space caching
> [29720.305743] btrfs: force clearing of disk cache
> 
> 
> I don't see any difference in behaviour with this mount option; it still 
> "hangs" for quite a bit at around ~1.8 GB (and reading with ~500 kB/s 
> when the hangs happens) on a subsequent dd run (several runs, 
> unmounts/mounts).

Right, you have to wait for all the caching threads to finish before you
unmount.

-chris