Hello, Any suggestion abort it?
whether you use which one, they both contact Cassandra via thrift. I'd suggest you to take look at RingCache, it can help you to compute the endpoint of data on Client side since Cassandra will forward your request on coordinator node. and another solution is that we may send data via the Gossiper Client, this can aviod the thrift overhead. Reference the process of bulk loader in Cassandra. Best, On Thu, Mar 21, 2013 at 3:27 PM, Alicia Leong <lccali...@gmail.com> wrote: > > I tried https://github.com/datastax/java-driver > with below CQL3. It works well. > > CREATE TABLE my_columnfamily ( > printer varchar, > computer varchar, > snapshot int, > xml text, > status varchar, > PRIMARY KEY ((printer, computer), snapshot) > ) > WITH CLUSTERING ORDER BY (snapshot DESC); > > CREATE INDEX my_columnfamily_status > > ON payloads (status); > > > Cluster cluster = > Cluster.builder().addContactPoint("localhost").withPort(9160).build(); > Session session = cluster.connect("my_keyspace"); > > for (Row row : session.execute("select * from my_columnfamily where status = > 'N'")) { > System.out.println("pinter :" + row.getString(0)); > System.out.println("computer :" + row.getString(1)); > System.out.println("snapshot :" + row.getInt(2)); > System.out.println("status :" + row.getString(3)); > System.out.println("xml :" + row.getString(4)); > System.out.println("------------------------------------------------"); > } > > > > > > > > On Thu, Mar 21, 2013 at 3:10 PM, Marko Asplund <marko.aspl...@gmail.com>wrote: > >> Víctor Hugo Oliveira Molinar wrote: >> > I guess Hector fits your requirements. The last release is pretty new. >> > But i'd suggest you to take a look at astyanax too. >> >> thanks. >> >> C* versions 1.1 and 1.2 seem to have been released after the latest >> Hector release. >> Does Hector support al the newer C* features (such as CQL3)? >> >> >> marko >> > > -- *Ric Dong *
