Steps:
1. Use debian-based image with k8s and docker elements
2. Add any Docker app to k8s cluster (Use default settings for k8s)
3. Deploy environment
Actual result:
Deployment fails with error:
2015-05-10 18:13:44 - [murano.engine.system.agent.AgentException]: {'errorCode': 100, 'message': u'Script setupFlannel returned error code', 'extra': None, 'details': {u'stdout': None, u'stderr': u'Error: cannot sync with the cluster using endpoints http://127.0.0.1:4001, http://127.0.0.1:2379', u'exitCode': 2}, 'time': u'2015-05-10 18:13:41.503504'}
2015-05-10 18:13:44 - Error: {'errorCode': 100, 'message': u'Script setupFlannel returned error code', 'extra': None, 'details': {u'stdout': None, u'stderr': u'Error: cannot sync with the cluster using endpoints http://127.0.0.1:4001, http://127.0.0.1:2379', u'exitCode': 2}, 'time': u'2015-05-10 18:13:43.318419'}
Trace in murano-agent:
.py:69
2015-05-10 18:13:43.317 555 DEBUG application [-] 'setupFlannelMaster.sh' execution stderr: 'Error: cannot sync with the cluster using endpoints http://127.0.0.1:4001, http://127.0.0.1:2379
' run /opt/stack/venvs/murano-agent/local/lib/python2.7/site-packages/muranoagent/executors/application/__init__.py:73
2015-05-10 18:13:43.317 555 ERROR muranoagent.app [-] Error running execution plan
2015-05-10 18:13:43.317 555 TRACE muranoagent.app Traceback (most recent call last):
2015-05-10 18:13:43.317 555 TRACE muranoagent.app File "/opt/stack/venvs/murano-agent/local/lib/python2.7/site-packages/muranoagent/app.py", line 89, in _run
2015-05-10 18:13:43.317 555 TRACE muranoagent.app result = runner.run()
2015-05-10 18:13:43.317 555 TRACE muranoagent.app File "/opt/stack/venvs/murano-agent/local/lib/python2.7/site-packages/muranoagent/execution_plan_runner.py", line 39, in run
2015-05-10 18:13:43.317 555 TRACE muranoagent.app raise script_globals['__execution_plan_exception']
2015-05-10 18:13:43.317 555 TRACE muranoagent.app CustomException: Script setupFlannel returned error code
2015-05-10 18:13:43.317 555 TRACE muranoagent.app
This happened, because command "/opt/bin/etcdctl mk /coreos.com/network/config '{"Network":"10.200.0.0/16"}'" fails with error:
Error: cannot sync with the cluster using endpoints http://127.0.0.1:4001, http://127.0.0.1:2379
This happened, because service etcd was dead.
root@murano-bbbsui9irwhai5-kube-1-plm3e4y6qhus:~# service etcd status
● etcd.service
Loaded: not-found (Reason: No such file or directory)
Active: inactive (dead)
We can use systemctl daemon-reload and get status of etcd one more time
root@murano-bbbsui9irwhai5-kube-1-plm3e4y6qhus:~# systemctl daemon-reload
root@murano-bbbsui9irwhai5-kube-1-plm3e4y6qhus:~# service etcd status
● etcd.service - LSB: Start distributed key/value pair service
Loaded: loaded (/etc/init.d/etcd)
Active: inactive (dead)
And start it:
root@murano-bbbsui9irwhai5-kube-1-plm3e4y6qhus:~# service etcd start
root@murano-bbbsui9irwhai5-kube-1-plm3e4y6qhus:~# service etcd status
● etcd.service - LSB: Start distributed key/value pair service
Loaded: loaded (/etc/init.d/etcd)
Active: active (exited) since Sun 2015-05-10 18:23:50 UTC; 2s ago
Process: 841 ExecStart=/etc/init.d/etcd start (code=exited, status=0/SUCCESS)
But nothing happened and process not running:
root@murano-bbbsui9irwhai5-kube-1-plm3e4y6qhus:~# ps aux | grep etcd
root 871 0.0 0.0 11120 1708 pts/0 R+ 18:25 0:00 grep etcd
And Flannel still can not sync.
root@murano-bbbsui9irwhai5-kube-1-plm3e4y6qhus:~# /opt/bin/etcdctl mk /coreos.com/network/config '{"Network":"10.200.0.0/16"}'
Error: cannot sync with the cluster using endpoints http://127.0.0.1:4001, http://127.0.0.1:2379
But if we use opts and binaries straight-forward we got working etcd service
http://paste.openstack.org/show/219143/ - etcd successful run
http://paste.openstack.org/show/219144/ - in other ssh executed ps aux | grep etcd
And after we up our services by hand, Flannel works fine:
http://paste.openstack.org/show/219149/
Is there link to the fix for this issue anywhere?