Re: [Zope] ZODB vs Relational Databases
It is encouraging to hear that the zope can handle some fairly large numbers. This was one of my biggest concerns. How is performance when retieving this data? Have you seen any degradation as the number of has grown, specifically have you seen any threshold that should be avoided? Thanks, Mike
I asked the same question not too long ago and the answer I got was to
give it
a try and see what it will handle. So far I have been impressed. It requires re-thinking if you are a RDBMS guy like me (and like you sound). For instance, do not store relations by id, or key, or other values. It'll slow things down, store them by path.
Say for instance you have a class and students. Each student can belong to a class....
/class/math /class/history
Those are the paths to the classes. Now in your student table, store that path. That is the relation.
Hope this helps. I have loaded it with 31102 * 5 objects. (5 Bible translations at 31102 verses (objects) per).
If you do decide to use RDBMS however, it plugs into zope very nicely, so no concerns their either.
Jeremy
_______________________________________________ Zope maillist - Zope@zope.org http://mail.zope.org/mailman/listinfo/zope ** No cross posts or HTML encoding! ** (Related lists - http://mail.zope.org/mailman/listinfo/zope-announce http://mail.zope.org/mailman/listinfo/zope-dev )
On Monday 10 March 2003 10:19 am, Michael Long wrote:
It is encouraging to hear that the zope can handle some fairly large numbers. This was one of my biggest concerns. How is performance when retieving this data? Have you seen any degradation as the number of has grown, specifically have you seen any threshold that should be avoided?
Mike, I am not certian of the answer to your question because I have not done extensive benchmarking yet. I can say that it searches the text of the Bible very quickly, but beyond that I'm not certain. I have not noticed visible changes in Zope's response, but as for raw numbers, I do not have any, sorry. Soon I will. I am still in the infancy of my project and am just now getting to a place where benchmarking would be even remotely useful. Maybe someone else has an idea? I have heard of ZODB being multiple g's, but second hand, not first. Jeremy
participants (2)
-
Jeremy Cowgar -
Michael Long