Hi, I'm running a 4 nodes Cassandra cluster, and I'm facing the same problem (node not present on nodetool ring, but unreachable on CLI describe cluster...). I'm currently running version 1.0.2, but I have update from 0.8.x, the problem may exist since a while, I don't really know. I can't stop my cluster, but I don't know how to apply the patch to repair the consequences of this bug by removing definitely this node.

Can someone tell me the way to proceed to apply a patch and call this java function ?

Thank you,


2011/11/18 Giannis Neokleous <giannis@generalsentiment.com>
Thanks for the info Brandon. I'll do the upgrade once 0.8.8 is released.

On Wed, Nov 16, 2011 at 2:43 PM, Brandon Williams <driftx@gmail.com> wrote:
On Mon, Nov 14, 2011 at 2:49 PM, Giannis Neokleous
<giannis@generalsentiment.com> wrote:
> Hello everyone,
> We're using the bulk loader to load data every day to Cassandra. The
> machines that use the bulkloader are diferent every day so their IP
> addresses change. When I do "describe cluster" i see all the unreachable
> nodes that keep piling up for the past few days. Is there a way to remove
> those IP addresses without terminating the whole cluster at the same time
> and restarting it?
> The unreachable nodes cause issues when we want to make schema changes to
> all the nodes or when we want to truncate a CF.
> Any suggestions?

It sounds like you're running into
https://issues.apache.org/jira/browse/CASSANDRA-3351 so the first step
would be to upgrade to a version that has it fixed.

Unfortunately, this won't solve the problem, just prevent it from
happening in the future.  To remove the old nodes, you can apply
https://issues.apache.org/jira/browse/CASSANDRA-3337 on one node and
call the JMX method for the unreachable endpoints.