bcache & btrfs: "Btrfs detected SSD devices, enabling SSD mode"

bcache & btrfs: "Btrfs detected SSD devices, enabling SSD mode"

[Rolf Fokkens]

This may be known already, but I noticed the following message in syslog 
during boot:

Btrfs detected SSD devices, enabling SSD mode

Btrfs is right in a sense: there is a SSD involved, but there's a HDD as 
well. I don't think it is an issue, but I have no idea how "btrfs' SSD 
optimizations"  will interact with bcache. Might be interesting.

Rolf

Re: bcache & btrfs: "Btrfs detected SSD devices, enabling SSD mode"

[Gabriel de Perthuis]

> This may be known already, but I noticed the following message in syslog
> during boot:
> 
> Btrfs detected SSD devices, enabling SSD mode
> 
> Btrfs is right in a sense: there is a SSD involved, but there's a HDD as
> well. I don't think it is an issue, but I have no idea how "btrfs' SSD
> optimizations"  will interact with bcache. Might be interesting.


From https://btrfs.wiki.kernel.org/ :
> SSD […] optimizations (e.g. avoiding unnecessary seek optimizations,
sending writes in clusters, even if they are from unrelated files. This
results in larger write operations and faster write throughput)

The optimisations should be beneficial iff bcache is used in writeback
mode.  ssd_spread and discard are enabled separately; I'm not sure what
ssd_spread (from the wiki: more strict about finding a large unused
region of the disk for new allocations) would do for bcache.

Re: bcache & btrfs: "Btrfs detected SSD devices, enabling SSD mode"

[Rolf Fokkens]

On 09/20/2013 06:25 PM, Gabriel de Perthuis wrote:
> SSD […] optimizations (e.g. avoiding unnecessary seek optimizations,
> sending writes in clusters, even if they are from unrelated files. This
> results in larger write operations and faster write throughput)
>
> The optimisations should be beneficial iff bcache is used in writeback
> mode.  ssd_spread and discard are enabled separately; I'm not sure what
> ssd_spread (from the wiki: more strict about finding a large unused
> region of the disk for new allocations) would do for bcache.
 From observing iostat I have the impression that bcache sometimes 
bypasses the cache, probably related to sequential_cutoff. Seems like 
btrfs tries to cluster writes (because an SSD is detected) and next 
bcache bypasses the SSD (because the writes are clustered). Doesn't seem 
the right kind of cooperation :-)

(Assuming of course that clustered writes are based on sequential writes)