Join HA cluster fails

Description

Setup 2 managers (m1, m2) running on CentOs.

Steps to reproduce:
1. Start HA cluster with the manager m1.
2. Switch to the manager m2
cfy profiles use...
3. Join HA cluster with the manager m2.

Expected result:
Manager m2 joined cluster successfully.

Actual result:
Join failed with following error: Timed out waiting for database replication to be established.

Attached log.

In addition, sometimes 'join' fails with 'Connection aborted' error.
For example log:
Joining the Cloudify Manager cluster: [u'10.0.0.5']
E 2018-02-15T08:33:17 Started /opt/manager/env/bin/create_cluster_node --config /tmp/tmpV9R8Vp.
E 2018-02-15T08:33:17 Starting /opt/manager/env/bin/create_cluster_node --config /tmp/tmpV9R8Vp...
E ('Connection aborted.', BadStatusLine("''",))
E This can happen when the manager is not working with SSL, but the client does

Assignee

Unassigned

Reporter

Dmitry Gertsman

Labels

None

Bug Type

regression bug

Target Version

4.3

Severity

Critical

Epic Link

Sprint

None

Fix versions

Affects versions

Configure