unit fails when changing rpc_interface config option
Bug #1713688 reported by
Gábor Mészáros
This bug affects 1 person
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Cassandra Juju Charm |
Fix Released
|
High
|
Stuart Bishop |
Bug Description
When changing the rpc interface to another network, the charm does not restart cassandra and fails to connect to the new interface:port, because cassandra is still listening on the old interface (but the configuration file gets rendered just fine).
Upon issueing `service cassandra restart`, the charm can continue and finishes the change.
Changed in cassandra-charm: | |
status: | New → Confirmed |
importance: | Undecided → High |
To post a comment you must log in.
The charm correctly noticed the IP address change and made a restart request to the leader. However, later in the hook the reset_auth_ keyspace_ replication handler failed, and it dies with a traceback (see the end of the attached charm.log). This is the real failure, which stopped the restart request being sent to the leader and should have put the unit into an error state. The reason it failed is most likely that the charm is attempting to contact the local cassandra server on the new interface, but because the restart hasn't happened yet the server is still listening on the old interface.
Are you able to confirm that the 100.70.0.134 IP address was the new interface?