I'm using CL=QUORUM (=Hector default) for both reads and writes. Most of
the times, the test passes, but sometimes it fails because I get back
the old value. Since the test is single-threaded, I guess it is a bug.
I'll try to reduce the test to something smaller that can be used for
troubleshooting.
By the way, is it documented somewhere under what circumstances one can
expect inconsistencies and when not?
On 7/19/2010 9:26 PM, Ran Tavory wrote:
if your test case is correct then it sounds like a bug to me. With one
node, unless you're writing with CL=0 you should get full consistency.
On Mon, Jul 19, 2010 at 10:14 PM, Hugo <h...@unitedgames.com
<mailto:h...@unitedgames.com>> wrote:
Hi,
Being fairly new to Cassandra I have a question on the eventual
consistency. I'm currently performing experiments with a
single-node Cassandra system and a single client. In some of my
tests I perform an update to an existing subcolumn in a row and
subsequently read it back from the same thread. More often than
not I get back the value I've written (and expected), but
sometimes it can occur that I get back the old value of the
subcolumn. Is this a bug or does it fall into the eventual
consistency?
I'm using Hector 0.6.0-14 on Cassandra 0.6.3 on a single disk,
double-core Windows machine with a Sun 1.6 JVM. All reads and
writes are quorum (the default), but I don't think this matters in
my setup.
Groets, Hugo.
