zfs discuss - Jul 2012 - Benefits of enabling compression in ZFS for the zones

Hi all,

By default I''m using ZFS for all the zones:

admjoresp at cyd-caszonesrv-15:~$ zfs list
NAME                         USED  AVAIL  REFER  MOUNTPOINT
opt                         4.77G  45.9G   285M  /opt
opt/zones                   4.49G  45.9G    29K  /opt/zones
opt/zones/glad-gm02-ftcl01   367M  45.9G   367M  /opt/zones/glad-gm02-ftcl01
opt/zones/glad-gp02-ftcl01   502M  45.9G   502M  /opt/zones/glad-gp02-ftcl01
opt/zones/glad-gp02-ftcl02  1.21G  45.9G  1.21G  /opt/zones/glad-gp02-ftcl02
opt/zones/mbd-tcasino-02     257M  45.9G   257M  /opt/zones/mbd-tcasino-02
opt/zones/mbd-tcasino-04     281M  45.9G   281M  /opt/zones/mbd-tcasino-04
opt/zones/mbfd-gp02-ftcl01   501M  45.9G   501M  /opt/zones/mbfd-gp02-ftcl01
opt/zones/mbfd-gp02-ftcl02   475M  45.9G   475M  /opt/zones/mbfd-gp02-ftcl02
opt/zones/mbhd-gp02-ftcl01   475M  45.9G   475M  /opt/zones/mbhd-gp02-ftcl01
opt/zones/mbhd-gp02-ftcl02   507M  45.9G   507M  /opt/zones/mbhd-gp02-ftcl02

However, I have the compression disabled in all of them.

According to this Oracle whitepaper 
http://www.oracle.com/technetwork/server-storage/solaris10/solaris-zfs-in-containers-wp-167903.pdf:

"The next example demonstrates the compression property. If compression 
is enabled, Oracle Solaris ZFS will transparently compress all of the 
data before it is written to disk. The benefits of compression
are both saved disk space and possible write speed improvements."

What exactly means "POSSIBLE" write speed improvements?

As you can see above I don''t use to have any room problems, so if
I''m
going to enable the compression flag it has to be because of the write 
speed improvements.

BTW, all the zones are containing Glassfish and they''re logging a lot. 
More than that, there''s not so much IO activity in the mentioned ZFS
fs.

-- 
I will face my fear. I will permit it to pass over me and through me.
And when it has gone past I will turn the inner eye to see its path.
Where the fear has gone there will be nothing. Only I will remain.

On Tue, Jul 10, 2012 at 4:25 PM, Jordi Espasa Clofent
<jespasac at minibofh.org> wrote:
> Hi all,
>
> By default I''m using ZFS for all the zones:
>
> admjoresp at cyd-caszonesrv-15:~$ zfs list
> NAME                         USED  AVAIL  REFER  MOUNTPOINT
> opt                         4.77G  45.9G   285M  /opt
> opt/zones                   4.49G  45.9G    29K  /opt/zones
> opt/zones/glad-gm02-ftcl01   367M  45.9G   367M 
/opt/zones/glad-gm02-ftcl01
> opt/zones/glad-gp02-ftcl01   502M  45.9G   502M 
/opt/zones/glad-gp02-ftcl01
> opt/zones/glad-gp02-ftcl02  1.21G  45.9G  1.21G 
/opt/zones/glad-gp02-ftcl02
> opt/zones/mbd-tcasino-02     257M  45.9G   257M  /opt/zones/mbd-tcasino-02
> opt/zones/mbd-tcasino-04     281M  45.9G   281M  /opt/zones/mbd-tcasino-04
> opt/zones/mbfd-gp02-ftcl01   501M  45.9G   501M 
/opt/zones/mbfd-gp02-ftcl01
> opt/zones/mbfd-gp02-ftcl02   475M  45.9G   475M 
/opt/zones/mbfd-gp02-ftcl02
> opt/zones/mbhd-gp02-ftcl01   475M  45.9G   475M 
/opt/zones/mbhd-gp02-ftcl01
> opt/zones/mbhd-gp02-ftcl02   507M  45.9G   507M 
/opt/zones/mbhd-gp02-ftcl02
>
> However, I have the compression disabled in all of them.
>
> According to this Oracle whitepaper
>
http://www.oracle.com/technetwork/server-storage/solaris10/solaris-zfs-in-containers-wp-167903.pdf:
>
> "The next example demonstrates the compression property. If
compression is
> enabled, Oracle Solaris ZFS will transparently compress all of the data
> before it is written to disk. The benefits of compression
> are both saved disk space and possible write speed improvements."
>
> What exactly means "POSSIBLE" write speed improvements?
compression = possibly less data  to write (depending on the data) possibly
faster writes

Some data is not compressible (e.g. mpeg4 movies), so in that case you
won''t see any improvements.

-- 
Fajar

On 2012-07-10 11:34, Fajar A. Nugraha wrote:
> compression = possibly less data  to write (depending on the data) >
possibly faster writes
>
> Some data is not compressible (e.g. mpeg4 movies), so in that case you
> won''t see any improvements.
Thanks for your answer Fajar.

As I said in my initial mail, those zones are mainly only writing some 
Glassfish logs. Since they all are text files, I guess I can save a lot 
of space using compression. Hopefully I can even improve the 
performance. Can I?

However. What''s the difference using

zfs set compression=on opt/zones/whatever_zone

or

zfs set compression=gzip-6 opt/zones/whatever_zone

or

zfs set compression=gzip-9 opt/zones/whatever_zone


TIA.


-- 
I will face my fear. I will permit it to pass over me and through me.
And when it has gone past I will turn the inner eye to see its path.
Where the fear has gone there will be nothing. Only I will remain.

On 07/10/12 09:25 PM, Jordi Espasa Clofent wrote:
> Hi all,
>
> By default I''m using ZFS for all the zones:
>
> admjoresp at cyd-caszonesrv-15:~$ zfs list
> NAME                         USED  AVAIL  REFER  MOUNTPOINT
> opt                         4.77G  45.9G   285M  /opt
> opt/zones                   4.49G  45.9G    29K  /opt/zones
> opt/zones/glad-gm02-ftcl01   367M  45.9G   367M 
/opt/zones/glad-gm02-ftcl01
> opt/zones/glad-gp02-ftcl01   502M  45.9G   502M 
/opt/zones/glad-gp02-ftcl01
> opt/zones/glad-gp02-ftcl02  1.21G  45.9G  1.21G 
/opt/zones/glad-gp02-ftcl02
> opt/zones/mbd-tcasino-02     257M  45.9G   257M  /opt/zones/mbd-tcasino-02
> opt/zones/mbd-tcasino-04     281M  45.9G   281M  /opt/zones/mbd-tcasino-04
> opt/zones/mbfd-gp02-ftcl01   501M  45.9G   501M 
/opt/zones/mbfd-gp02-ftcl01
> opt/zones/mbfd-gp02-ftcl02   475M  45.9G   475M 
/opt/zones/mbfd-gp02-ftcl02
> opt/zones/mbhd-gp02-ftcl01   475M  45.9G   475M 
/opt/zones/mbhd-gp02-ftcl01
> opt/zones/mbhd-gp02-ftcl02   507M  45.9G   507M 
/opt/zones/mbhd-gp02-ftcl02
>
> However, I have the compression disabled in all of them.
>
> According to this Oracle whitepaper
>
http://www.oracle.com/technetwork/server-storage/solaris10/solaris-zfs-in-containers-wp-167903.pdf:
>
> "The next example demonstrates the compression property. If
compression
> is enabled, Oracle Solaris ZFS will transparently compress all of the
> data before it is written to disk. The benefits of compression
> are both saved disk space and possible write speed improvements."
>
> What exactly means "POSSIBLE" write speed improvements?
With compression enabled, less data has to be written to disk, so N 
bytes writes in N/compress ratio time.

On most systems, the performance cost of compressing and uncompressing 
data is relatively low.
> As you can see above I don''t use to have any room problems, so if
I''m
> going to enable the compression flag it has to be because of the write
> speed improvements.
I always enable compression by default and only turn it off for 
filesystems I know hold un-compressible data such as media files.

-- 
Ian.

On Tue, Jul 10, 2012 at 4:40 PM, Jordi Espasa Clofent
<jespasac at minibofh.org> wrote:
> On 2012-07-10 11:34, Fajar A. Nugraha wrote:
>
>> compression = possibly less data  to write (depending on the data)
>> possibly faster writes
>>
>> Some data is not compressible (e.g. mpeg4 movies), so in that case you
>> won''t see any improvements.
>
>
> Thanks for your answer Fajar.
>
> As I said in my initial mail, those zones are mainly only writing some
> Glassfish logs. Since they all are text files, I guess I can save a lot of
> space using compression. Hopefully I can even improve the performance. Can
> I?
>
correct.

Even "normal" OS files are usually compressible-enough. For example,
this is my root partition (Ubuntu, but uses zfs nontheles)

$ sudo zfs get compression,compressratio C/ROOT/precise
NAME            PROPERTY       VALUE     SOURCE
C/ROOT/precise  compression    gzip      inherited from C
C/ROOT/precise  compressratio  2.48x     -

so on that dataset I save over 50% I/O read/writes (in bytes)
> However. What''s the difference using
>
> zfs set compression=on opt/zones/whatever_zone
on = standard LZJB compression (very fast, but doesn''t compress much)
>
> or
>
> zfs set compression=gzip-6 opt/zones/whatever_zone
gzip-6 and gzip uses gzip compression. Fast enough, good compression.
>
> or
>
> zfs set compression=gzip-9 opt/zones/whatever_zone
gzip-9 = uses gzip''s best, but also slowest, compression

-- 
Fajar

Thanks for you explanation Fajar. However, take a look on the next lines:

# available ZFS in the system

root at sct-caszonesrv-07:~# zfs list
NAME                         USED  AVAIL  REFER  MOUNTPOINT
opt                          532M  34.7G   290M  /opt
opt/zones                    243M  34.7G    32K  /opt/zones
opt/zones/sct-scw02-shared   243M  34.7G   243M  /opt/zones/sct-scw02-shared
static                       104K  58.6G    34K  /var/www/

# creating a file in /root (UFS)

root at sct-caszonesrv-07:~# dd if=/dev/zero of=file.bin count=1024 bs=1024
1024+0 records in
1024+0 records out
1048576 bytes (1.0 MB) copied, 0.0545957 s, 19.2 MB/s
root at sct-caszonesrv-07:~# pwd
/root

# enable compression in some ZFS zone

root at sct-caszonesrv-07:~# zfs set compression=on opt/zones/sct-scw02-shared

# copying the previos file to this zone

root at sct-caszonesrv-07:~# cp /root/file.bin 
/opt/zones/sct-scw02-shared/root/

# checking the file size in the origin dir (UFS) and the destination one 
(ZFS with compression enabled)

root at sct-caszonesrv-07:~# ls -lh /root/file.bin
-rw-r--r-- 1 root root 1.0M Jul 10 13:21 /root/file.bin

root at sct-caszonesrv-07:~# ls -lh /opt/zones/sct-scw02-shared/root/file.bin
-rw-r--r-- 1 root root 1.0M Jul 10 13:22 
/opt/zones/sct-scw02-shared/root/file.bin

# the both files has exactly the same cksum!

root at sct-caszonesrv-07:~# cksum /root/file.bin
3018728591 1048576 /root/file.bin

root at sct-caszonesrv-07:~# cksum /opt/zones/sct-scw02-shared/root/file.bin
3018728591 1048576 /opt/zones/sct-scw02-shared/root/file.bin

So... I don''t see any size variation with this test.
Actually I tried the same with compression=gzip/gzip-6 or even gzip-9 
and is always the same. I understand it shouldn''t be?

What am I missing or doing wrong?

More info about the system:

root at sct-caszonesrv-07:~# cat /etc/release
                     Oracle Solaris 10 8/11 s10x_u10wos_17b X86
   Copyright (c) 1983, 2011, Oracle and/or its affiliates. All rights 
reserved.
                             Assembled 23 August 2011


-- 
I will face my fear. I will permit it to pass over me and through me.
And when it has gone past I will turn the inner eye to see its path.
Where the fear has gone there will be nothing. Only I will remain.

Of course you don''t see any difference, this is how it should work.
''ls'' will never report the compressed size, because
it''s not aware of it.
Nothing is aware of the compression and decompression that takes place
on-the-fly, except of course zfs.

That''s the reason why you could gain in write and read speed if you use
compression, because the actual amount of compressed data that is being
written and read from the pool is smaller than the original data. And I
think with the checksum test you prooved that zfs checksums the
uncompressed data.

Regards,
Feci

On Tue, Jul 10, 2012 at 1:29 PM, Jordi Espasa Clofent <jespasac at
minibofh.org
> wrote:
> Thanks for you explanation Fajar. However, take a look on the next lines:
>
> # available ZFS in the system
>
> root at sct-caszonesrv-07:~# zfs list
>
> NAME                         USED  AVAIL  REFER  MOUNTPOINT
> opt                          532M  34.7G   290M  /opt
> opt/zones                    243M  34.7G    32K  /opt/zones
> opt/zones/sct-scw02-shared   243M  34.7G   243M
>  /opt/zones/sct-scw02-shared
> static                       104K  58.6G    34K  /var/www/
>
> # creating a file in /root (UFS)
>
> root at sct-caszonesrv-07:~# dd if=/dev/zero of=file.bin count=1024 bs=1024
> 1024+0 records in
> 1024+0 records out
> 1048576 bytes (1.0 MB) copied, 0.0545957 s, 19.2 MB/s
> root at sct-caszonesrv-07:~# pwd
> /root
>
> # enable compression in some ZFS zone
>
> root at sct-caszonesrv-07:~# zfs set compression=on
> opt/zones/sct-scw02-shared
>
> # copying the previos file to this zone
>
> root at sct-caszonesrv-07:~# cp /root/file.bin
/opt/zones/sct-scw02-shared/**
> root/
>
> # checking the file size in the origin dir (UFS) and the destination one
> (ZFS with compression enabled)
>
> root at sct-caszonesrv-07:~# ls -lh /root/file.bin
> -rw-r--r-- 1 root root 1.0M Jul 10 13:21 /root/file.bin
>
> root at sct-caszonesrv-07:~# ls -lh /opt/zones/sct-scw02-shared/**
> root/file.bin
> -rw-r--r-- 1 root root 1.0M Jul 10 13:22 /opt/zones/sct-scw02-shared/**
> root/file.bin
>
> # the both files has exactly the same cksum!
>
> root at sct-caszonesrv-07:~# cksum /root/file.bin
> 3018728591 1048576 /root/file.bin
>
> root at sct-caszonesrv-07:~# cksum /opt/zones/sct-scw02-shared/**
> root/file.bin
> 3018728591 1048576 /opt/zones/sct-scw02-shared/**root/file.bin
>
> So... I don''t see any size variation with this test.
> Actually I tried the same with compression=gzip/gzip-6 or even gzip-9 and
> is always the same. I understand it shouldn''t be?
>
> What am I missing or doing wrong?
>
> More info about the system:
>
> root at sct-caszonesrv-07:~# cat /etc/release
>                     Oracle Solaris 10 8/11 s10x_u10wos_17b X86
>   Copyright (c) 1983, 2011, Oracle and/or its affiliates. All rights
> reserved.
>                             Assembled 23 August 2011
>
>
>
> --
> I will face my fear. I will permit it to pass over me and through me.
> And when it has gone past I will turn the inner eye to see its path.
> Where the fear has gone there will be nothing. Only I will remain.
>  ______________________________**_________________
> zfs-discuss mailing list
> zfs-discuss at opensolaris.org
>
http://mail.opensolaris.org/**mailman/listinfo/zfs-discuss<http://mail.opensolaris.org/mailman/listinfo/zfs-discuss>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20120710/7ef322bd/attachment.html>

> From: zfs-discuss-bounces at opensolaris.org [mailto:zfs-discuss-
> bounces at opensolaris.org] On Behalf Of Jordi Espasa Clofent
> 
> root at sct-caszonesrv-07:~# zfs set compression=on opt/zones/sct-scw02-
> shared
If you use compression=on, or lzjb, then you''re using very fast
compression.
Should not hurt performance, in fact, may gain performance for highly
compressible data.

If you use compression=gzip (or any gzip level 1 thru 9) then you''re
using a
fairly expensive compression algorithm.  It will almost certainly hurt
performance, but you may gain more disk space. (Probably not.)

> # copying the previos file to this zone
> 
> ...
> # checking the file size in the origin dir (UFS) and the destination one
> (ZFS with compression enabled)
> 
> ...
> 
> # the both files has exactly the same cksum!
When you enable filesystem compression, it is all done behind the scenes.
It is intended to look and behave exactly the same at the file level, so you
should not see any difference in your files.  Your files should appear to be
exactly the same, and have exactly the same size and checksum as you
otherwise would have expected.  It would be bad if you had to explain to
your users, "Hey you guys, I''ve enabled filesystem compression, so
you
should expect all your files to have different sizes and different
checksums."  The compression & decompression happen at a low level - So
all
your applications don''t know about it.  

If you want to see the difference in disk usage, I believe "df" and
"du"
will report different things...  Or perhaps "ls -l" versus
"du" ...

On 07/10/12 12:45, Ferenc-Levente Juhos wrote:
> Of course you don''t see any difference, this is how it should
work.
> ''ls'' will never report the compressed size, because
it''s not aware of
> it. Nothing is aware of the compression and decompression that takes
> place on-the-fly, except of course zfs.
> That''s the reason why you could gain in write and read speed if
you use
> compression, because the actual amount of compressed data that is being
> written and read from the pool is smaller than the original data. And I
> think with the checksum test you prooved that zfs checksums the
> uncompressed data.
No ZFS checksums are over the data as it is stored on disk so the 
compressed data.

-- 
Darren J Moffat

On 2012-07-10 13:45, Ferenc-Levente Juhos wrote:
> Of course you don''t see any difference, this is how it should
work.
> ''ls'' will never report the compressed size, because
it''s not aware of
> it. Nothing is aware of the compression and decompression that takes
> place on-the-fly, except of course zfs.
Yep, my fault. You are right:
root at sct-caszonesrv-07:/opt/test# du -hs u*
1.4M    unix1
820K    unix2


Good article in here:
https://blogs.oracle.com/observatory/entry/zfs_compression_a_win_win
> That''s the reason why you could gain in write and read speed if
you use
> compression, because the actual amount of compressed data that is being
> written and read from the pool is smaller than the original data. And I
> think with the checksum test you prooved that zfs checksums the
> uncompressed data.
Thanks for the explanation.
-- 
I will face my fear. I will permit it to pass over me and through me.
And when it has gone past I will turn the inner eye to see its path.
Where the fear has gone there will be nothing. Only I will remain.

On Tue, Jul 10, 2012 at 6:29 AM, Jordi Espasa Clofent
<jespasac at minibofh.org> wrote:
> Thanks for you explanation Fajar. However, take a look on the next lines:
>
> # available ZFS in the system
>
> root at sct-caszonesrv-07:~# zfs list
>
> NAME                         USED  AVAIL  REFER  MOUNTPOINT
> opt                          532M  34.7G   290M  /opt
> opt/zones                    243M  34.7G    32K  /opt/zones
> opt/zones/sct-scw02-shared   243M  34.7G   243M 
/opt/zones/sct-scw02-shared
> static                       104K  58.6G    34K  /var/www/
>
> # creating a file in /root (UFS)
>
> root at sct-caszonesrv-07:~# dd if=/dev/zero of=file.bin count=1024 bs=1024
> 1024+0 records in
> 1024+0 records out
> 1048576 bytes (1.0 MB) copied, 0.0545957 s, 19.2 MB/s
> root at sct-caszonesrv-07:~# pwd
> /root
>
> # enable compression in some ZFS zone
>
> root at sct-caszonesrv-07:~# zfs set compression=on
opt/zones/sct-scw02-shared
>
> # copying the previos file to this zone
>
> root at sct-caszonesrv-07:~# cp /root/file.bin
> /opt/zones/sct-scw02-shared/root/
>
> # checking the file size in the origin dir (UFS) and the destination one
> (ZFS with compression enabled)
>
> root at sct-caszonesrv-07:~# ls -lh /root/file.bin
> -rw-r--r-- 1 root root 1.0M Jul 10 13:21 /root/file.bin
>
> root at sct-caszonesrv-07:~# ls -lh
/opt/zones/sct-scw02-shared/root/file.bin
> -rw-r--r-- 1 root root 1.0M Jul 10 13:22
> /opt/zones/sct-scw02-shared/root/file.bin
>
> # the both files has exactly the same cksum!
>
> root at sct-caszonesrv-07:~# cksum /root/file.bin
> 3018728591 1048576 /root/file.bin
>
> root at sct-caszonesrv-07:~# cksum
/opt/zones/sct-scw02-shared/root/file.bin
> 3018728591 1048576 /opt/zones/sct-scw02-shared/root/file.bin
>
> So... I don''t see any size variation with this test.
ls(1) tells you how much data is in the file - that is, how many bytes
of data that an application will see if it reads the whole file.
du(1) tells you how many disk blocks are used.  If you look at the
stat structure in stat(2), ls reports st_size, du reports st_blocks.

Blocks full of zeros are special to zfs compression - it recognizes
them and stores no data.  Thus, a file that contains only zeros will
only require enough space to hold the file metadata.

$ zfs list -o compression ./
COMPRESS
      on

$ dd if=/dev/zero of=1gig count=1024 bs=1024k
1024+0 records in
1024+0 records out

$ ls -l 1gig
-rw-r--r--   1 mgerdts  staff    1073741824 Jul 10 07:52 1gig

$ du -k 1gig
0	1gig

-- 
Mike Gerdts
http://mgerdts.blogspot.com/

2012-07-10 15:49, Edward Ned Harvey wrote:
> If you use compression=on, or lzjb, then you''re using very fast
compression.
> Should not hurt performance, in fact, may gain performance for highly
> compressible data.
>
> If you use compression=gzip (or any gzip level 1 thru 9) then
you''re using a
> fairly expensive compression algorithm.  It will almost certainly hurt
> performance, but you may gain more disk space. (Probably not.)
Well, as far as the discussion relates to zones, the "WORM"
(write once - read many) type of data, such as the OS image
of the local zone, can "suffer" gzip-9 compression during
installation of the zone and applications. This may make
your files consume less disk sectors and further reads can
be faster. Then you can enable lzjb on the same dataset,
and further writes (of logs) would be compressed faster.

In fact, you might want to "delegate" a dataset to the
zone and create several filesystems in it, with different
compression options, for your logs, application data and
perhaps databases (which may be sensitive to IO block size
and dislike external compression in favor of speeds). For
these there is a "zle" compression which only compresses
blocks filled with zeroes - that allows to save some space
when your DB precreates a huge storage file but only uses
a few kilobytes in it.

I am not qualified to state whether gzip decompression might
be slower during reads than lzjb or not, but remember that
all this relies on general assumption that current CPUs are
overqualified for their jobs and have lots of spare cycles -
so (de)compression has little impact on real work anyway.
Also decompression tends to be faster than compression,
because there is little to no analysis to do - only matching
compressed tags to a dictionary of original data snippets.

HTH,
//Jim Klimov

To amplify what Mike says...

On Jul 10, 2012, at 5:54 AM, Mike Gerdts wrote:
> ls(1) tells you how much data is in the file - that is, how many bytes
> of data that an application will see if it reads the whole file.
> du(1) tells you how many disk blocks are used.  If you look at the
> stat structure in stat(2), ls reports st_size, du reports st_blocks.
> 
> Blocks full of zeros are special to zfs compression - it recognizes
> them and stores no data.  Thus, a file that contains only zeros will
> only require enough space to hold the file metadata.
> 
> $ zfs list -o compression ./
> COMPRESS
>      on
> 
> $ dd if=/dev/zero of=1gig count=1024 bs=1024k
> 1024+0 records in
> 1024+0 records out
> 
> $ ls -l 1gig
> -rw-r--r--   1 mgerdts  staff    1073741824 Jul 10 07:52 1gig
"ls -ls" shows the length (as in -l) and size (as in -s,
units="blocks")
So you can see that it takes only space for metadata.
   1 -rw-r--r--   1 root     root     1073741824 Nov 26 06:52 1gig
size ................................ length


 -- richard

--
ZFS Performance and Training
Richard.Elling at RichardElling.com
+1-760-896-4422







-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20120710/fdcababe/attachment.html>