I ran across this awhile back and I mentioned "partitioning for speed" in another post and I think it's worth mentioning here.  It's not about Sesame per se but it has to do with maximizing speed, which is often a goal with databases.

Robin Harris on ZDNet mentioned awhile back that companies with high transaction rates often specially partition their drives.  The basic concept has to do with angular velocity.  I mentioned to him that I have Win 7 on an old Western Digital 80GB drive.  He said that even if the electronics are identical, larger drives will have faster data transfer rates.  Modern hard disks are designed that they automatically store data on the outer tracks first.  In that position, at a given spindle speed, for a given period of time a greater amount of data will pass the read/write head.  He said that for that reason some companies deliberately partition large drives and only store data on the outer (first) partition.  Also, with the greater data density of larger drives, the increase in transfer rate can be substantial, even, e.g., going from a 300GB drive with 50GB used to a 2TB drive.

(I have a 2-drive system with each as a wholly-independent bootable drive.  I rarely use Win 7.)  My main drive is a 1.5TB partitioned as 100GB C: and the rest as a data partition.  Keeping the first partition small forces everything on it to the fastest part of the disk.

(And when SSD's drop a bit more in price, I will be able to easily replace the 100GB partition with a 64GB SSD.)