Hi Alex,
I made full restore of tpcc.fbk with loging of each compresion request.

Result is here:

http://www.elektlabs.cz/tpcc.7z

Record size is original length of request for compresion.

New RLE is RLE that I was developed for FB3, but was rejected.

LZ4 is fast version, not HC, just original data packed only by lz4.

RLE+LZ4 is new RLE and over this result is runned LZ4.

Some notes:
1. tpcc database has big differencies to live DB.
     the biggest record has Average unpacked length: 689.00, compression 
ratio: 1.22
     In UTF8 era one VARCHAR(100) occupied 400 bytes.

2. I don't test speed impact, full integration of lz4 will cause changes 
in vio.cpp and dmp.epp.
   But I still belive, that we have tu put some thashold, by current 
test  it can be around 256 bytes of RLE output.

3. Because encoding and compresion are based on statistical probabilities,
    we need some real data for research.

If you are interesting in record encoding/compresion, I'm ready to help.

Slavek

Ing. Slavomir Skopalik
Executive Head
Elekt Labs s.r.o.
Collection and evaluation of data from machines and laboratories
by means of system MASA (http://www.elektlabs.cz/m2demo)
-----------------------------------------------------------------
Address:
Elekt Labs s.r.o.
Chaloupky 158
783 72 Velky Tynec
Czech Republic
---------------------------------------------------------------
Mobile: +420 724 207 851
icq:199 118 333
skype:skopaliks
e-mail:skopa...@elektlabs.cz
http://www.elektlabs.cz

On 17.3.2015 15:24, Alex Peshkoff wrote:
> On 03/17/15 16:50, Slavomir Skopalik wrote:
>> Hi Alex,
>> please can you take your tpcc.fdb,  backup, 7zip and  send to me?
>> I will use as reference database.
>>
> Done.
> Confirm that you've received it pls.
>
>
> ------------------------------------------------------------------------------
> Dive into the World of Parallel Programming The Go Parallel Website, sponsored
> by Intel and developed in partnership with Slashdot Media, is your hub for all
> things parallel software development, from weekly thought leadership blogs to
> news, videos, case studies, tutorials and more. Take a look and join the
> conversation now. http://goparallel.sourceforge.net/
> Firebird-Devel mailing list, web interface at 
> https://lists.sourceforge.net/lists/listinfo/firebird-devel
>



------------------------------------------------------------------------------
Dive into the World of Parallel Programming The Go Parallel Website, sponsored
by Intel and developed in partnership with Slashdot Media, is your hub for all
things parallel software development, from weekly thought leadership blogs to
news, videos, case studies, tutorials and more. Take a look and join the 
conversation now. http://goparallel.sourceforge.net/
Firebird-Devel mailing list, web interface at 
https://lists.sourceforge.net/lists/listinfo/firebird-devel

Reply via email to