On Jan 25, 2015, at 8:41 PM, John Sanda <jsanda(a)redhat.com>
wrote:
If an error occurred while adding one of the nodes and the replication_factor failed to
get updated, then running repair won’t help. You need to log into the storage node cluster
using cqlsh to check the replication_factor with the following command,
describe keyspace rhq;
It should set to 2.
I made a mistake. The command should be,
describe keyspace system_auth;
And its replication_factor should be 2. You should check the rhq keyspace as well as both
should be 2. The command to change it is,
alter keyspace system_auth with replication = {‘replication_factor’: 2};
After making the change run repair first on the system_auth keyspace as you did
originally. Then run repair on the rhq keyspace. Hope that helps.
- John