You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@cassandra.apache.org by Felipe Esteves <fe...@b2wdigital.com> on 2016/11/11 18:22:52 UTC

User auth error during new DC rebuild

Hi, I've got a strange behaviour when adding a new DC to my Cassandra
cluster.

Cassandra version is 2.1.8.

The topology is a 3 DC cluster (let's call it A, B, C)  and I have 3 users
(1 superuser, 2 non-superuser)
I'm adding a new DC (let's call it D) with 1 node, so I started it with
auto_bootstrap=false, changed the keyspaces replication to include the new
DC and ran "nodetool rebuild C" from the D node.

I've altered the application keyspace as well the system_auth replication.

However, 1 of the non-superusers is getting authorization error when I
connect to any node on C (the DC that is streaming the rebuild) or D (the
new DC).
If I connect to A or B, it works.
The superuser and the other non-superuser connects in all nodes.


Any tips on why this is happening? Doesn't make sense to me just one of the
users not working.

The error is:

code=0100 [Bad credentials] message="Username and/or password are incorrect"



Felipe Esteves

Tecnologia

felipe.esteves@b2wdigital.com <se...@b2wdigital.com>

--