[ckan-dev] SQLite removal?

Fri Jun 14 21:41:53 UTC 2013

On Fri, Jun 14, 2013 at 6:49 PM, Ian Ward <ian at excess.org> wrote:

> On Fri, Jun 14, 2013 at 10:59 AM, David Read
> <david.read at hackneyworkshop.com> wrote:
> > The time taken on postgres is nothing to do with CKAN core or the
> > specific tests. It is a result of postgres dropping dozens of tables
> > and recreating them. I think when David did the current postgres test
> > setup it was in haste and it is clear to anyone reading logging during
> > tests that it could be improved somewhat. So please do improve it! :-)
>
> I hacked up my own tests to use a template database instead of
> creating from scratch each time.  It takes access to a more privileged
> postgres user, but it's really worth it.
>

That sounds interesting.  A few developers put the postgres cluster on a
ramdisc so that may be more problematic.   However having the tables there
and not recreating them at test start sounds feasible and was made more
difficult by the fact that sqlite memory db needs to be created at the
start and we did not want the code paths to vary too much.

Also having a dbdump with the create test data was an option which would
make lots of tests faster, but that was postgres specific too.

>
> _______________________________________________
> ckan-dev mailing list
> ckan-dev at lists.okfn.org
> http://lists.okfn.org/mailman/listinfo/ckan-dev
> Unsubscribe: http://lists.okfn.org/mailman/options/ckan-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.okfn.org/pipermail/ckan-dev/attachments/20130614/faa23eed/attachment-0001.html>