ZFS/compression/performance

[Dennis Glatting]

On Wed, 2011-10-12 at 08:59 +0100, Steven Hartland wrote:
----- Original Message -----
> From: "Dennis Glatting" <freebsd at penx.com>
>
>
> >I would appreciate someone knowledgeable in ZFS point me in the right
> > direction.
> >
> > I have several ZFS arrays, some using gzip for compression. The
> > compressed arrays hold very large text documents (10MB->20TB)
> > and are highly compressible. Reading the files from a compressed
> > data sets is fast with little load. However, writing to the
> > compressed data sets incurs substantial load on the order of a
> >  load average from 12 to 20.
> >
> > My questions are:
> >
> > 1) Why such a heavy load on writing?
> > 2) What kind  of limiters can I put into effect to reduce load
> >    without impacting compressibilty? For example, is there some
> >    variable to controls the number of parallel compression
> >    operations?
> >
> > I have a number of different systems. Memory is 24GB on each of the 
two
> > large data systems, SSD (Revo) for cache, and a SATA II ZIL. One 
system is
> > a 6 core i7 @ 3.33 GHz and the other 4 core ii7 @ 2.93 GHz. The arrays 
are
> > RAIDz using cheap 2TB disks.
>
> Have you tried using the alternative compression algorithms
> e.g. lzjb or gzip-[1-5] the default gzip = gzip-6
>

I have tried lzjb and I am unimpressed. I have not tried different levels 
of gzip on ZFS but I have tried it on documents with results I expected.

As I mentioned, I have a lot of data. Two files were 26GB uncompressed but 
I had to kill those data sets because I ran out of room (I have 
reorganized my arrays since then). My ZFS compression ratio is 4.93x and I 
would require more storage at different gzip levels or ljzb.

An option is not too compress with ZFS rather directly with gzip however I 
would still need lots of temporary storage for manipulation, which is what 
I am doing now (e.g., sort). Processing with zcat isn't always a good 
solution because some applications want files, but you have to do what you 
have to do.



>     Regards
>     Steve
>
> ================================================
> This e.mail is private and confidential between Multiplay (UK) Ltd. and 
the person or entity to whom it is addressed. In the event of 
misdirection, the recipient is prohibited from using, copying, printing or 
otherwise disseminating it or any information contained in it.
>
> In the event of misdirection, illegible or incomplete transmission 
please telephone +44 845 868 1337
> or return the E.mail to postmaster at multiplay.co.uk.
>
> _______________________________________________
> freebsd-fs at freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-fs
> To unsubscribe, send any mail to "freebsd-fs-unsubscribe at freebsd.org"
>
-- 
Dennis Glatting <dg at pki2.com>