Hi. I just got the job to find out how we (the technical university Graz) could build up a large image-database for different users.. Zope came to my mind because of its oject oriented structure. And because I'd really like to do Zope work at work :). I now have to find out how Zope can be used with really large databases. (100000 to 1mio objects theoretically in the future). Can anyone provide me with URLs or information on: The relational Database Storage Manager (couldnt find any references) (is this intented for large databases?) Or splitting the Zope hirarchy into multiple ZODB FileStorage Manager instances. SO that I could e.g. save on subfolder of the root folder in on data.fs and the next subfolder in another data.fs I am thinking of the filesize limit of some operating systems. 2GB isnt really much. I think that a FileStorage would be better for our content (images), because performance issues (never saved 20GB in an relational database before :) ) Any help is appreciated, Mario Lang <mlang@home.delysid.org> Technical University Graz Department Computing Services Phone +43-316-873-8508 E-Mail: lang@zid.tu-graz.ac.at
I think that a FileStorage would be better for our content (images), because performance issues (never saved 20GB in an relational database before :) )
I am using Zope for an large project at the moment. For static content, such as images, I use apache to server them up (and will probably move to something faster such as thttpd or kernel mode http server) and then just store a URL reference to the actual data. This leaves zope to do presenatation, which it does really well and let a more suitable server handle the static content. Any way, I'm not saying that zope couldn't handle this however depending on your exact use this may be an option for you. Cheers, Benno
participants (2)
-
Ben Leslie -
Mario Lang