Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
Home -> Community -> Usenet -> c.d.o.server -> Re: expdp question
On Aug 16, 2:07 am, Brian Peasland <d..._at_nospam.peasland.net> wrote:
> Chuck wrote:
> > We're finally getting around to installing 10g and I had a question
> > about expdp. We used to write our exports to named pipes so they could
> > be compressed with gzip. Data pump doesn't support that. Is there *any*
> > way around that limitation or do I need to continue using export/import
> > if I want compressed files?
If absolutely *any* way to compress *on a fly*, then you may want to
try something along these lines:
Start expdp:
- specify DUMPFILE with one or more names containing substitution
variable(s).
- specify FILESIZE
Watch when a dumpfile is finished. This is a tricky bit; dumpfiles are opened and closed lots of times and it's difficult to determine when they're finally closed for good. Just waiting when fuser returns nothing won't help you. I would take a look at status command in interactive-command interface ('size' and 'bytes written'); just remember that the last few dumpfiles are smaller than FILESIZE.
When dumpfile is finally closed, start gzipping it.
This should work with parallel workers as well because AFAIK only one worker writes to a particular dumpfile.
All-in-all I'd rather get enough disk space and gzip after expdp is finished.
> > TIA
>
> Upgrade to 11g? It does allow for compression of data pump dump files.
It's a good joke :) The following heated exchange just proves this.
Regards,
Igor
> Cheers,
> Brian
>
> --
> ===================================================================
>
> Brian Peasland
> d...@nospam.peasland.nethttp://www.peasland.net
>
> Remove the "nospam." from the email address to email me.
>
> "I can give it to you cheap, quick, and good.
> Now pick two out of the three" - Unknown
>
> --
> Posted via a free Usenet account fromhttp://www.teranews.com
Received on Tue Aug 21 2007 - 03:52:12 CDT