Thank you Ran, that's working! Should have tried that as well. There is a difference in the log's, embedded shows: INFO 20:16:31,796 Binding thrift service to localhost/127.0.0.1:9160
Batch-file-started one: INFO thrift.CassandraDaemon: Listening for thrift clients... So maybe different classes are being used for thrift commication based on how the server is started. As soon as I figure that out I'll post it. On Fri, Aug 27, 2010 at 7:57 PM, Ran Tavory <ran...@gmail.com> wrote: > did you try connecting to a real cassandra instance, not an embedded one? > I use an embedded one for testing and it works, but just to narrow down your > problem. > > On Fri, Aug 27, 2010 at 6:13 PM, Ruben de Laat <ru...@logic-labs.nl> wrote: >> >> Hi, >> >> I am new to cassandra, so maybe I am missing something obvious... >> Version: Latest nightly build (2010-08-23_13-57-40), but same results >> with 0.7.0b1 >> >> Server code (default configuration file): >> >> System.setProperty("cassandra.config", "conf/cassandra.yaml"); >> EmbeddedCassandraService embeddedCassandraService = new >> EmbeddedCassandraService(); >> embeddedCassandraService.init(); >> >> Client code: >> >> Socket socket = new Socket("127.0.0.1", 9160); >> TSocket transport = new TSocket(socket); >> TBinaryProtocol tBinaryProtocol = new TBinaryProtocol(transport); >> Client client = new Client(tBinaryProtocol); >> System.out.println(client.describe_cluster_name()); >> >> The problem is that it hangs/blocks on the >> "client.describe_cluster_name()" call, actually it hangs on any call I >> have tried. >> I was first trying with the Pelops client, but that one is using the >> Thrift API as well, so this is narrowed down. >> >> I have already tried multiple different combination of creating the >> client (different transports). >> I have also tried with "thrift_framed_transport_size_in_mb: 0" >> disabling framed transports. >> >> Starting the client without a running server gives a proper >> "Connection refused", so some sort of connection is definitely made. >> >> Thanks and Kind regards, >> Ruben > >