durusmail: durus-users: Re: [Schevo-devel] scalability of schevo
Re: [Schevo-devel] scalability of schevo
Re: [Schevo-devel] scalability of schevo
2006-03-01
2006-03-01
2006-03-01
Re: [Schevo-devel] scalability of schevo
David Binger
2006-03-01
On Mar 1, 2006, at 4:59 AM, mario ruggier wrote:

> On Mar 1, 2006, at 2:46 AM, David Binger wrote:
>> The idea is to write the index in some format other than
>> pickle that can be used without loading the whole thing
>> first.
>
> This implies a separate file for the db's index? Which seems like a
> good idea to me, considering also that this index is anyway derived
> data... And, if you want to regenerate the index, just remove the
> file and restart the db.

No, the packed index is currently written into the same file as the
rest of the data, and this is important because it forces the
index and the object records to stay synchronized.

Yes the index derived data, but deriving it requires reading through
the entire
data file, and that can take too long.





reply