Craig Ringer wrote:
> On 19/04/2010 4:51 PM, Craig Ringer wrote:
>
>> There might be another possible approach that uses the system
>> "xmin/xmax" fields of each tuple. That'd permit your incremental dumps
>> to be done read-only, saving you a whole lot of expensive I/O and bloat.
>> I'm just not
On 19/04/2010 4:51 PM, Craig Ringer wrote:
There might be another possible approach that uses the system
"xmin/xmax" fields of each tuple. That'd permit your incremental dumps
to be done read-only, saving you a whole lot of expensive I/O and bloat.
I'm just not sure what I'm thinking of will wor
Ostrovsky Eugene wrote:
> Hi.
> I need to export data from the database to external file. The difficulty
> is that only data modified or added since previous export should be
> written to the file.
> I consider adding "modification_time" timestamp field to all the tables
> that should be exported.
Hi.
I need to export data from the database to external file. The difficulty is
that only data modified or added since previous export should be written to the
file.
I consider adding "modification_time" timestamp field to all the
tables that should be exported. Then I can set this field to now()
