[ZODB-Dev] High Write Applications

[Phillip J. Eby]

At 11:59 AM 8/4/03 -0400, Casey Duncan wrote:
>  about a seemingly straightforward way to improve
>write performance for FileStorage. It was inspired by bsddb which allows
>something similar.
>
>You would divide the filestorage into two files, a read-only file for past
>transactions and a write file for new transactions. At first, you would have
>only the write file. When you pack, the packed transactions would go into the
>read file and a new write file would be started. The index could keep track
>of the file boundary so that, given an oid, the storage could figure out
>which file the current data resides in. These files would be put on different
>drive spindles to improve performance.

I'm not clear on how that would help with the primary issue with supporting 
high write volumes: ZODB transactions write a *lot* more data than an RDBMS 
would for an equivalent transaction.  If there wasn't so much data being 
written, the issues your approach is addressing would have less 
impact.  But, they are only a small part of the overall problem, as least 
as I understand it.