just to clarify the previous proposal. you don't specify which mount point a file belongs to. you specify which znodes represent a container entry point (what you are calling a mount point). a znode will be created in the partition (aka container) that the subtree belongs to. i believe in this sense the two proposals are both the same.
On Thu, Jun 9, 2011 at 5:10 PM, Alexander Shraer <[email protected]> wrote: > This is a preliminary proposal, so everything is still open. Still, I think > there are many advantages over the previous namespace partitioning proposal > (http://wiki.apache.org/hadoop/ZooKeeper/PartitionedZookeeper) that wasn't > implemented AFAIK. The idea here is to make much smaller and more intuitive > changes. > For example, the previous proposal did not offer any ordering guarantees > across partitions. Also - in Linux mount you don't need to specify for each > new file which mount point the file belongs to - we can exploit the tree > structure to infer that instead of creating and maintaining an additional > hierarchy like in the previous proposal. >
