Mike, You are correct, there should only be one instance of Stanchion. Stanchion's role is to serve as a serialization point for requests that create or modify system entities that need to be globally unique [1]. e.g. Buckets or user accounts. These are entities where there is not enough context available for that application-level conflict resolution to function in a way that would maintain a good user experience. For example, if two users were allowed to create the same bucket, there is not really a good way to determine who should get to own it and the process of informing the loser could be expensive and disruptive, especially if they were allowed to begin loading data into that bucket. One thing we do to try and avoid such a scenario is to serialize all bucket creation requests through Stanchion. At this time it is a single application and this does represent a single point of failure, but only for system operations involving creation or modification of buckets or users. Object access is unaffected if Stanchion is not running. Still, we realize this is less than ideal and there is work being done now to let multiple nodes to work together to eliminate this as a single point of failure.
[1] : http://docs.basho.com/riakcs/latest/cookbooks/Globally-Unique-Entities/ Kelly On Wed, Jun 5, 2013 at 5:22 PM, Mike O'Toole <mike.oto...@outlook.com>wrote: > Im doing some analysis on object stores for an upcoming project and I am > prototyping Riak-CS right now. Im a little confused about Stanchions > role in all this. Is there really only supposed to be one stanchion > instance per cluster? > > Regards, Mike > > _______________________________________________ > riak-users mailing list > riak-users@lists.basho.com > http://lists.basho.com/mailman/listinfo/riak-users_lists.basho.com > >
_______________________________________________ riak-users mailing list riak-users@lists.basho.com http://lists.basho.com/mailman/listinfo/riak-users_lists.basho.com
