Thread: Incremental Backup

Resp. Sir /Madam,
      We are hosting a website who is having PostgreSQL RDBMS as back-end. In that we have to upload data onto the
serverfrom the standalone application on a daily basis. At present, on Standalone application side we are taking the
backupof the whole database, While on the server side we are deleting, droping, creation of new tables and inserting
thedata into it. This is fine untill the data size is small, but it is increasing in abundant.  
So is there any way of doing some kind of incremental backup in which only the daily work which is done is backed up
fromthe standalone machine and is appended into the data which is already existing over there on the server. 
The reply to this mail should be considered as an emergency case and i hope you cooperation for it.

Thanks,
Regards,

The Guardian.

On Tue, Aug 19, 2003 at 07:19:19PM -0500, theguardian@vsnl.net wrote:
> So is there any way of doing some kind of incremental backup in
> which only the daily work which is done is backed up from the
> standalone machine and is appended into the data which is already
> existing over there on the server.

Not at the moment.  You could use a replication system to solve this
for you, though.  See contrib/rserv, the recent improvements to rserv
available on gborg (I haven't used them), contrib/dbmirror, or the
soon-to-be-released _other_ improvements to contrib/rserv.

A
--
----
Andrew Sullivan                         204-4141 Yonge Street
Liberty RMS                           Toronto, Ontario Canada
<andrew@libertyrms.info>                              M2P 2A8
                                         +1 416 646 3304 x110

On Tue, 19 Aug 2003 theguardian@vsnl.net wrote:

> Resp. Sir /Madam,
>       We are hosting a website who is having PostgreSQL RDBMS as back-end. In that we have to upload data onto the
serverfrom the standalone application on a daily basis. At present, on Standalone application side we are taking the
backupof the whole database, While on the server side we are deleting, droping, creation of new tables and inserting
thedata into it. This is fine untill the data size is small, but it is increasing in abundant.  
> So is there any way of doing some kind of incremental backup in which only the daily work which is done is backed up
fromthe standalone machine and is appended into the data which is already existing over there on the server. 
> The reply to this mail should be considered as an emergency case and i hope you cooperation for it.

There are a few ways to do this.

One is to have a seperate server that only has today's data, it gets
backed up, the data gets pushed to live (if need be) and it gets wiped
clean.  Or use schemas and have one for each day that you can then backup
each day.