deploy fails at rabbitmq because epmd does not start
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
kolla-ansible |
Expired
|
Undecided
|
Unassigned |
Bug Description
This is related to #1855935, but I was asked to create a separate bugreport.
I am running kolla-ansible:
+ exec /usr/sbin/
econnrefused
Protocol 'inet_tcp': register/listen error:
It turned out my hosts file had duplicate entries, since kolla-ansible simply appends to /etc/hosts, and I had already prepared it fully on all hosts.
But I have now tried with several versions of hosts files, and it never changes anything, even with a single "172.16.102.100 ctrl1.lab2.stack" entry.
--Original hosts file--
[root@compute1 ~(admin)]# cat /etc/hosts
127.0.0.1 localhost
::1 localhost localhost.
172.16.102.100 ctrl1.lab2.stack controller1.
172.16.102.109 ctrl2.lab2.stack controller2.
172.16.102.101 compute1.lab2.stack compute1
172.16.102.102 compute2.lab2.stack compute2
172.16.102.103 compute3.lab2.stack compute3
172.16.102.104 neutron1.lab2.stack neutron1
172.16.102.105 neutron2.lab2.stack neutron2
172.16.102.106 storage1.lab2.stack storage1
172.16.102.107 storage2.lab2.stack storage2
172.16.102.108 storage3.lab2.stack storage3
# BEGIN ANSIBLE GENERATED HOSTS
172.16.102.100 ctrl1.lab2.stack ctrl1
172.16.102.104 neutron1.lab2.stack neutron1
172.16.102.101 compute1.lab2.stack compute1
172.16.102.106 storage1.lab2.stack storage1
# END ANSIBLE GENERATED HOSTS
--End
I also tried switching to the "kolla/
What I have noticed just now though, since the econnrefused doesnt sound like rev-lookup failure, is that starting a shell in kolla_toolbox container, and manually running epmd in that one, will then make easy to start the rabbitmq server in its own container.
What I cant figure out, is how epmd is supposed to be started "normally".
I've tried upping the log-level of rabbitmq with "log.default.level = debug", but it doesn't seem to kick in, and I'm quite a n00b on rabbitmq.
RabbitMQ spawns its own epmd for itself in the same container. Your issue is indeed different than the other's. Your /etc/hosts is just fine, same duplicates are fine. You can turn off kolla-ansible creating new by disabling hosts management but it is irrelevant in here.
Did you tweak anything related to networking in comparison to standard install?