Offline indexing issues

Ruby Apps Ferret
1

If I disable ferret in my environment file and then run a cron job
every hour to index the records that have been added/changed, do I
enable Ferret just in the script that does the hourly indexing? Or do
I somehow need to tell the process that runs the Ferret DRB server
that it should start writing to the index again? Thanks in advance.

Erik

2

Follow Up. I have 5 mongrels and a ferret drb server running on the
same machine. In environment.rb I disable ferret like so:

MyModel.disable_ferret

I have a script that runs every hour that does the following:

MyModel.enable_ferret
MyModel.bulk_index all_records_created_or_changed_in_the_last_hour

Now, ClassMethods defines bulk_index as:

 def bulk_index(*ids)
   ids = ids.first if ids.size == 1 && ids.first.is_a?(Enumerable)
   aaf_index.bulk_index(ids)
 end

aaf_index is an instance of RemoteIndex that dispatches bulk_index to
the drb server, correct? If that’s the case, then the Server class
sends the method call to the aaf_index, which has bulk_index defined
in ferret_extensions.rb, correct? the bulk_index method uses the
BulkIndexer.index_records method, which checks to see if ferret is
enbled for each record: rec.ferret_enabled?(true). Now, if this all
happens in the Drb server, wouldn’t ferret be disabled there as it is
obviously using the same environment.rb file that disables ferret for
MyModel?

Thanks in advance.

3

On Thu, Oct 04, 2007 at 11:56:36AM -0400, Erik M. wrote:

Now, ClassMethods defines bulk_index as:
BulkIndexer.index_records method, which checks to see if ferret is
enbled for each record: rec.ferret_enabled?(true). Now, if this all
happens in the Drb server, wouldn’t ferret be disabled there as it is
obviously using the same environment.rb file that disables ferret for
MyModel?

Exactly. Since with bulk_index (and rebuild_index, too) the
ferret_enabled? check always only happens in the DRb server, calling
MyModel.enable_ferret on the client side has no effect.

I just fixed this - instance.ferret_enabled? now ignores the
class level ferret_enabled state when called by the bulk indexer (which
is also used by the rebuild_index method, so the same behaviour applies
there).

I think the behaviour now is more useful - when you disable ferret at
the class level and then explicitly call bulk_index or rebuild_index on
this class, aaf assumes you have a good reason to do so and just does
what you say.

Cheers,
Jens

–
Jens Krämer
webit! Gesellschaft für neue Medien mbH
Schnorrstraße 76 | 01069 Dresden
Telefon +49 351 46766-0 | Telefax +49 351 46766-66
[email protected] | www.webit.de

Amtsgericht Dresden | HRB 15422
GF Sven Haubold, Hagen Malessa