I rather doubt anyone has tested any of these timings. I've not seen
anything published, anyway. I suggest running them yourself to see
what's fastest.
I have noticed that (at least on FreeBSD), a seperate bzip2/gzip doesn't
utilize a second CPU, which is odd. But it's been a long time since I've
looked at this kind of thing.
If you do run tests, please share your findings with the community
(pgsql-general or -performance might be more appropriate than -admin).
On Wed, Oct 12, 2005 at 09:36:07AM -0400, Chris Hoover wrote:
> I'm wondering, does anyone know of any timing tests for pg_dump?
>
> I am trying to find the fastest way to do our nightly database backups using
> pg_dump. And I just wanted to make sure I don't reinvent the wheel if
> someone has already done a bunch of comparisions with the various ways to
> backup.
>
> Here are the kind of timings I'm looking at:
>
> pg_dump -Fc -Zx
> pg_dump -Fc
> pg_dump -Ft
> pg_dump
> pg_dump -Fc | gzip > backup
> pg_dump -Ft | gzip > backup
> pg_dump | gzip > backup
> pg_dump -Fc | bzip2 > backup
> pg_dump -Ft | bzip2 > backup
> pg_dump | bzip2 > backup
>
> I know there are many variables, but I'm just looking at the pg_dump
> process. Does anyone know what the fastest pg_dump would be?
>
> Also, what -Z compression level equals the compression of gzip? I've played
> with -Z9, but it seems to compress better than gzip (but takes a bit
> longer).
>
> Thanks,
>
> Chris
--
Jim C. Nasby, Sr. Engineering Consultant jnasby@pervasive.com
Pervasive Software http://pervasive.com work: 512-231-6117
vcard: http://jim.nasby.net/pervasive.vcf cell: 512-569-9461