On 31.10.2008 22:17 Uhr, Christophe Combelles wrote:
Tim Cook a écrit :
Does anyone know of any work being done on a PyTables
No, but I might be interested in such a thing, since I've created last year a
medical data analysis software on which I encountered such memory issues, while
looping over large structures in zodb.
I have looked at pytables a few times, but I've never tried it yet. (I had also
looked at SOOM (a component of NetEpi Analysis), which allow to compute stats on
very large datasets.)
Is it worth creating a new storage implementation for such cases? Don't just we
need to implement (or use an existing) specific structure in filestorage?
I see a future need in my project to be able to manage real time storage
of medical data and the author of PyTables lists the memory usage of the
ZODB as being a drawback in these situations.
The problem with all such implementations is: they don't support
transactions which is a major requirement. I also looked at 10gen lately
(something similar like GAE but open-source) and they also provide some
nice storage (of course "the cloud") but also lack transaction integration.
org:ZOPYX Ltd. & Co. KG
For more information about ZODB, see the ZODB Wiki:
ZODB-Dev mailing list - ZODB-Dev@zope.org