Thread

  1. Feature Request: bzip2 support in pg_dump, pg_restore

    Daniel Migowski <dmigowski@ikoffice.de> — 2010-08-13T08:57:13Z

    Hi, dear postgres developers,
    
    A small investigation showed to me that bzip2 compressed sql files take 
    only 60% of the space of gz compressed files. Since bzip2 is fairly 
    common today, could one add an option to pg_dump and pg_restore 
    supporting this compression type in their custom format? Or do the 
    PostgreSQL folks prefer no to do that regarding backwards compatibility?
    
    Regards,
    Daniel Migowski
    
    -- 
    *I**K*Office
    *UNTERNEHMENSSOFTWARE*
    	IKOffice GmbH 	*Daniel Migowski*
    
    	Mail: 	dmigowski@ikoffice.de <mailto:dmigowski@ikoffice.de>
    Nordstrasse 10 	Tel.: 	+49 (0)441 21 98 89 52
    26135 Oldenburg 	Fax.: 	+49 (0)441 21 98 89 55
    http://www.ikoffice.de 	Mob.: 	+49 (0)176 22 31 20 76
    
    Geschäftsführer: 	Ingo Kuhlmann, Daniel Migowski
    Amtsgericht Oldenburg: 	HRB 201467
    Steuernummer: 	64/211/01864
    
    
  2. Re: Feature Request: bzip2 support in pg_dump, pg_restore

    Adrian von Bidder <avbidder@fortytwo.ch> — 2010-08-13T09:31:32Z

    Heyho!
    
    On Friday 13 August 2010 10.57:13 Daniel Migowski wrote:
    > A small investigation showed to me that bzip2 compressed sql files take 
    > only 60% of the space of gz compressed files.
    
    But bzip2 is very slow.  I think if there should be changes to the data 
    compression, xz is probably the future.  xz is said to be similar to bzip2 
    in terms of compression ratio, but much faster.
    
    cheers
    -- vbi
    
    -- 
    And those meanies used what we said to you and they misconstrued it to
    mean what we said, and that's so unfair.
            -- SCO lawyers about IBM (paraphrased - groklaw)
    
  3. Re: Feature Request: bzip2 support in pg_dump, pg_restore

    Scott Ribe <scott_ribe@killerbytes.com> — 2010-08-13T13:50:49Z

    pg_dumpall | bzip2 > mydump.txt.bz2
    
    bunzip2 -kc mydump.txt.bz2 | bin/psql template1
    
    
    -- 
    Scott Ribe
    scott_ribe@elevated-dev.com
    http://www.elevated-dev.com/
    (303) 722-0567 voice
    
    
    
    
    
    
  4. Re: Feature Request: bzip2 support in pg_dump, pg_restore

    Greg Smith <greg@2ndquadrant.com> — 2010-08-13T13:57:35Z

    Daniel Migowski wrote:
    > A small investigation showed to me that bzip2 compressed sql files 
    > take only 60% of the space of gz compressed files. Since bzip2 is 
    > fairly common today, could one add an option to pg_dump and pg_restore 
    > supporting this compression type in their custom format? Or do the 
    > PostgreSQL folks prefer no to do that regarding backwards compatibility?
    
    http://wiki.postgresql.org/wiki/Developer_FAQ#Why_aren.27t_there_more_compression_options_when_dumping_tables.3F
    now has a first round answer to this topic, one that addresses the 
    things that pop up every time this discussion is raised.  There are 
    speed, code quality, code license, and patent issues all making this 
    much harder to change than is obvious at first.
    
    -- 
    Greg Smith  2ndQuadrant US  Baltimore, MD
    PostgreSQL Training, Services and Support
    greg@2ndQuadrant.com   www.2ndQuadrant.us
    
    
  5. Re: Feature Request: bzip2 support in pg_dump, pg_restore

    Tom Lane <tgl@sss.pgh.pa.us> — 2010-08-13T14:47:30Z

    Greg Smith <greg@2ndquadrant.com> writes:
    > Daniel Migowski wrote:
    >> A small investigation showed to me that bzip2 compressed sql files 
    >> take only 60% of the space of gz compressed files. Since bzip2 is 
    >> fairly common today, could one add an option to pg_dump and pg_restore 
    >> supporting this compression type in their custom format? Or do the 
    >> PostgreSQL folks prefer no to do that regarding backwards compatibility?
    
    > http://wiki.postgresql.org/wiki/Developer_FAQ#Why_aren.27t_there_more_compression_options_when_dumping_tables.3F
    > now has a first round answer to this topic, one that addresses the 
    > things that pop up every time this discussion is raised.  There are 
    > speed, code quality, code license, and patent issues all making this 
    > much harder to change than is obvious at first.
    
    Looks good.  I added the point that pg_dump output has to be readable
    far into the future, so conservatism in what we support is a good thing.
    
    			regards, tom lane