I found a fix for this one, rather a workaround.I changed the rpc_server_type in cassandra.yaml, from hsha to sync, and the error went away. I guess, there is some issue with the thrift nonblocking server.ThanksGurpreetOn Wed, May 16, 2012 at 7:04 PM, Gurpreet Singh <firstname.lastname@example.org> wrote:
Thanks Aaron. will do!On Mon, May 14, 2012 at 1:14 PM, aaron morton <email@example.com> wrote:
Are you using framed transport on the client side ?Try the Hector user list for hector specific help https://groups.google.com/forum/?fromgroups#!searchin/hector-usersCheersOn 12/05/2012, at 5:44 AM, Gurpreet Singh wrote:This is hampering our testing of cassandra a lot, and our move to cassandra 1.0.9.Has anyone seen this before? Should I be trying a different version of cassandra?
/GOn Thu, May 10, 2012 at 11:29 PM, Gurpreet Singh <firstname.lastname@example.org> wrote:
Hi,i have created 1 node cluster of cassandra 1.0.9. I am setting this up for testing reads/writes.I am seeing the following error in the server system.logERROR [Selector-Thread-7] 2012-05-10 22:44:02,607 TNonblockingServer.java (line 467) Read an invalid frame size of 0. Are you using TFramedTransport on the client side?Initially i was using a old hector 0.7.x, but even after switching to hector 1.0-5 and thrift version 0.6.1, i still see this error.I am using 20 threads writing/reading from cassandra. The max write batch size is 10 with payload size constant per key to be 600 bytes.On the client side, i see Hector exceptions happenning coinciding with these messages on the server.Any ideas why these errors are happenning?ThanksGurpreet