Hi Chap, If you have the ability to test the 2.0.0rc2 driver, I would recommend doing so, even from a dedicated test client or a JUnit test case. There are other benefits to the change, such as being able to use BatchStatements, aside from possible impact on your read timeouts.
Steve On Wed, Feb 5, 2014 at 3:06 PM, Chap Lovejoy <chaplove...@gmail.com> wrote: > Hi Steve, > > Thanks for the reply. After all that information in my initial message I > would forget one of the most important bits. We're running Cassandra 2.0.3 > with the 1.0.4 version of the DataStax driver. I'd seen mention of those > timeouts under earlier 2.x versions and really hoped they were the source > of our problem but unfortunately that doesn't seem to be the case. > > Thanks again, > Chap > > > On 5 Feb 2014, at 17:49, Steven A Robenalt wrote: > > Hi Chap, >> >> You don't indicate which version of Cassandra and what client side driver >> you are using, but I have seen the same behavior with Cassandra 2.0.2 and >> earlier versions of the Java Driver. With Cassandra 2.0.3 and the 2.0.0rc2 >> driver, my read timeouts are basically nonexistent at my current load >> levels. >> >> Not sure how this applies if you're still on 1.x versions of Cassandra >> since we moved off of that branch a few months ago. Ditto for the client >> driver if you're using something other than the Java Driver, or the 1.x >> version of same. Our problems were due to changes specific to the 2.x >> versions only. >> >> Steve >> > -- Steve Robenalt Software Architect HighWire | Stanford University 425 Broadway St, Redwood City, CA 94063 srobe...@stanford.edu http://highwire.stanford.edu
