[Zope] BTreeFolder for "huge" ZODB (was : populating a "huge" ZODB)

[Oliver Marx]

If I understand you right then you want to use ZODB as th=E9 DB for an
application?

\Oliver

>>-----Original Message-----
>>From: zope-admin@zope.org [mailto:zope-admin@zope.org]On Behalf Of
>>Pierre Godefroy
>>Sent: 21. juni 2002 12:37
>>To: zope@zope.org
>>Subject: [Zope] BTreeFolder for "huge" ZODB (was : populating a "huge"
>>ZODB)
>>
>>
>>Thank you very much Igor, however I am still a bit afraid because I see=
 a
>>scrolling window for the products contrained by the BTreeFolder on the
>>product homepage
>>(http://www.zope.org/Members/hathawsh/BTreeFolder). In the
>>exmple given (4096 items), it could still be OK, but I think having
>>1,000,000 + objects might be a problem. There is an interface
>>problem here
>>: the contents of such a folder can only be reasonably displayed in
>>successive batches. (in fact it is also another aspect of the problem o=
f
>>having thousands of objects in an ordinary folder : if you click on the
>>folder to open it, the Management interface will hang)
>>
>>Pierre
>>
>>
>>>On Fre, 2002-06-21 at 11:57, Pierre Godefroy wrote:
>>>[ huge folder ]
>>>Have a look at BTreeFolder, the whole structure is stored in a btree, =
so
>>>a lookup for an id should be pretty fast...
>>>greetings,
>>>Igor
>>
>>--
>>>Hi everybody,
>>>
>>>How would you populate a "huge" ZODB with hundreds of thousands
>>of objects
>>>which have no specific reason to be hierarchically distributed in a
>>>complex tree of folders and sub-folders?
>>>
>>>They could in fact all be in the same folder. It seems to be OK for
>>>hundreds of objects, but afterwards there is an incremental cost : eac=
h
>>>time you add a new object, there is a check on the existence of an
>>>identical ID before adding/creating it. When you reach the level of
>>>already existing thousands/tens of thousands/hundreds of
>>thousands objects
>>>in the same folder it seems you are in serious trouble...
>>>
>>>I know the question is a bit naive, and that usually you have a data
>>>structure which dictates a naturally "hierarchy" of folders which mean=
s
>>>that no folder would contain maybe more than a few hundred objects or
>>>sub-folders.
>>>
>>>Maybe I should "artificially" design an ID structure which would
>>>correspond to a hierarchy : for instance ojects would be identified by
>>>numbers with 7 digits, each digit corresponding to a lower level in th=
e
>>>hierarchy? It means that an object with ID 2343789 would be placed in
>>>the  "/2/3/4/3/7/8/9" sub-folder.
>>>
>>>What do you think? Any other idea?
>>
>>
>>
>>
>>
>>Pierre Godefroy
>>52 rue des Archives
>>75004 PARIS
>>FRANCE
>>
>>T=E9l. : +33 (0)1 42 74 46 05
>>
>>
>>
>>_______________________________________________
>>Zope maillist  -  Zope@zope.org
>>http://lists.zope.org/mailman/listinfo/zope
>>**   No cross posts or HTML encoding!  **
>>(Related lists -
>> http://lists.zope.org/mailman/listinfo/zope-announce
>> http://lists.zope.org/mailman/listinfo/zope-dev )
>>---
>>Incoming mail is certified Virus Free.
>>Checked by AVG anti-virus system (http://www.grisoft.com).
>>Version: 6.0.371 / Virus Database: 206 - Release Date: 13-06-2002
>>
---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.371 / Virus Database: 206 - Release Date: 13-06-2002