Hi, I have worked with Memcached and Redis. I might be able to add some value here. Can I have a high level overview of nature of data and sizing. Cassandra will be a complex setup from deployment point of view, so its better to rule that out. I have not tried Hadoop so can't comment on that. You can however use AVRO (hadoop project) serialization/deserialization and store a complete JSON payload against a key into one of the above memory datastructures (both for redis/memcached) and have tested with 600 writes/sec and 1000 reads/sec.
Thanks & Regards, Devesh On Sun, Feb 6, 2011 at 10:20 AM, Umashanthi Pavalanathan < [email protected]> wrote: > On Sun, Feb 6, 2011 at 10:16 AM, Luciano Resende <[email protected] > >wrote: > > > On Sat, Feb 5, 2011 at 12:40 PM, Umashanthi Pavalanathan > > <[email protected]> > > > > > > What is the back-end storage mechanism expected for the future > releases? > > Is > > > it the JCR mechanism? > > > I can have a look at it and come up with the storage structure of > > activities > > > etc. > > > > > > > As for the storage system, we currently use a JCR repository, but I > > believe this is something we can discuss. If you think that a > > different storage mechanism would work better with the relationships > > and other requirements from open social, please start a new thread and > > let's discuss this with the community. Personally, I've been working > > with some noSQL databases, and think they can be another option for > > us. > > > > Sure. I'm not much familiar with the JCR repository. I'll first look into > it > and then will post my opinions. > > > Thanks, > ~Umashanthi > > > > > > -- > > Luciano Resende > > http://people.apache.org/~lresende > > http://twitter.com/lresende1975 > > http://lresende.blogspot.com/ > > >
