[ome-users] pgpool-II with OMERO?

[Josh Moore]

On Apr 25, 2012, at 3:48 PM, Harri Jäälinoja wrote:
> 
> Hi all,

Hi Harri,

> we started to think that it would be nice to make a high-availability set up for the PostgrSQL database that OMERO is using, to make sure that no data is lost. Is anyone out there doing something like that? The first tool I found is pgpool-II, is anybody using that?

I don't want to dissuade anyone else out there from responding, but we'd suggest first you look at the newer versions of PG and it's built in log shipping, e.g.

  http://www.postgresql.org/docs/9.1/static/hot-standby.html

Cheers,
~Josh

> My initial guess about database activity in OMERO is that it is mostly just reading from the DB. There will be writing in the DB when users import new data, but if something goes wrong at that point, it will be easy to redo the import, so nothing is really lost. Then there are times when users add analysis results or metadata. Analysis results probably are generated by running some scripts, so those, too, should be easy to redo. Manually made annotations are another issue, it would be nasty to loose those. But here we could then balance the issue between what the performance overhead of the replication system is and how many hours of annotation work the user can loose.
> 
> Any comments, does this sound reasonable? Any other points to consider?
> 
> Best regards,
> Harri