HAproxy binds to IPv6 even if it's disabled in the OS and as a result fails to start
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Charm Helpers |
Fix Committed
|
Medium
|
Tytus Kurek | ||
OpenStack Base Layer |
Fix Released
|
Medium
|
Tytus Kurek | ||
charms.openstack |
Fix Released
|
Medium
|
Tytus Kurek |
Bug Description
This issue has been faced when deploying designate charm from openstack-
$ juju status designate
Model Controller Cloud/Region Version SLA
default bratislava bratislava 2.2.2 unsupported
App Version Status Scale Charm Store Rev OS Notes
designate 2.0.0 blocked 3 designate local 33 ubuntu
designate-hacluster active 3 hacluster local 31 ubuntu
Unit Workload Agent Machine Public address Ports Message
designate/51* blocked idle 22/lxd/23 100.107.2.1 9001/tcp Ports which should be open, but are not: 9001
designate-
designate/53 blocked idle 23/lxd/23 100.107.66.239 9001/tcp Ports which should be open, but are not: 9001
designate-
designate/54 blocked idle 24/lxd/23 100.107.66.242 9001/tcp Ports which should be open, but are not: 9001
designate-
Machine State DNS Inst id Series AZ Message
22 started 100.107.2.134 q8eh4f trusty FC1 Deployed
22/lxd/23 started 100.107.2.1 juju-64cbea-
23 started 100.107.66.126 fbg4ta trusty FC2 Deployed
23/lxd/23 started 100.107.66.239 juju-64cbea-
24 started 100.107.66.128 gdpgys trusty FC2 Deployed
24/lxd/23 started 100.107.66.242 juju-64cbea-
Relation Provides Consumes Type
cluster designate designate peer
dns-backend designate designate-bind regular
ha designate designate-hacluster subordinate
identity-service designate keystone regular
shared-db designate percona-cluster regular
amqp designate rabbitmq-
hanode designate-hacluster designate-hacluster peer
This is because "/etc/haproxy/
global
log 127.0.0.1 local0
log 127.0.0.1 local1 notice
maxconn 20000
user haproxy
group haproxy
spread-checks 0
defaults
log global
mode tcp
option tcplog
option dontlognull
retries 3
timeout queue 5000
timeout connect 5000
timeout client 30000
timeout server 30000
listen stats
mode http
stats enable
stats hide-version
stats realm Haproxy\ Statistics
stats uri /
stats auth admin:oetycPlNC
frontend tcp-in_
bind *:9001
bind :::9001
acl net_100.107.2.1 dst 100.107.
use_backend designate-
default_backend designate-
backend designate-
balance leastconn
server designate-51 100.107.2.1:8991 check
server designate-53 100.107.66.239:8991 check
server designate-54 100.107.66.242:8991 check
The problem is that "bind :::9001" statement won't work on the systems with IPv6 disabled and thus, will result in haproxy failing to start. It looks like the base openstack layer does not contain a check for this statement in "haproxy.cfg" template file.
tags: | added: 4010 cpe-onsite |
Changed in charms.openstack: | |
status: | Triaged → In Progress |
Changed in layer-openstack: | |
status: | Triaged → In Progress |
Changed in charm-helpers: | |
status: | Triaged → In Progress |
Changed in charm-helpers: | |
status: | In Progress → Fix Committed |
Changed in layer-openstack: | |
status: | In Progress → Fix Committed |
Changed in layer-openstack: | |
status: | Fix Committed → Fix Released |
Fix proposed to branch: master /review. openstack. org/507033/
Review: https:/