Re: Geographical redundancy

OK, well accepting data loss (even if it is "just" 6-12 hours worth)
really opens up a lot of possibilities...... EXCEPT that you also said you
want both sites to be able to modify data. Again, there is no real
multi-master replication available for postgres, so you'll have to have
both sites at least write to the same database server.

On Tue, 2 Jan 2007, Dennis wrote:

> Well, I am mainly concerned with catastrophic failure. If 1st (main)
> datacenter fails majorly (say fire, earthquake, db server dies etc), I
> need to be able to restore websites/data quickly in another location. If
> I get a data loss of say 6-12 hours during a major failure (which should
> never occur), I am ok with that.
>
> Ben <bench(at)silentmedia(dot)com> wrote: On Sat, 30 Dec 2006, Dennis wrote:
>
>> I was thinking of maybe just having 2nd location receive a PG dump (full
>> or incremental) every so often (an hour to 6 hours) and if the main
>> location fails majorly, restore the PG cluster from the dump and switch
>> DNS settings on the actual sites. I can make sure all website files are
>> always in sync on both locations.
>
> Well, first off, you can just rsync your archived WAL files. That may be
> easier than playing with pg_dump:
>
> http://www.postgresql.org/docs/8.2/interactive/continuous-archiving.html
>
> But second, and more important given your data-loss desires, if you do it
> this way you have a window where you can experience data loss.
> Specifically, after a transaction is committed, that commit will be at
> risk until the next transfer has completed.
>
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam? Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com