Description of problem: Trigger 3.6 OCP cluster, it resulted in that NO_PROXY was set mistakenly in /etc/sysconfig/docker, actually we didn't set proxy related variables in inventory hosts file. Version-Release number of the following components: openshift-ansible-3.6.136-1.git.0.ac6bb62.el7.noarch.rpm" How reproducible: always Steps to Reproduce: 1.Trigger 3.6 installation without proxy variables set Actual results: # egrep '^HTTP_PROXY|^HTTPS_PROXY|^NO_PROXY' /etc/sysconfig/docker /etc/sysconfig/atomic-openshift-node /etc/sysconfig/docker:NO_PROXY='.cluster.local,.svc,qe-cl-auto-dgwp-node-zone1-primary-1' Expected results: We shouldn't set the no_proxy if we don't set it in inventory host file. Additional info: <--snip--> TASK [openshift_facts : Gather Cluster facts and set is_containerized if needed] *** Thursday 06 July 2017 15:39:49 +0000 (0:00:00.835) 0:00:29.254 ********* changed: [qe-cl-auto-dgwp-master-etcd-zone2-1.0706-5vc.qe.rhcloud.com] => { "ansible_facts": { "openshift": { "common": { "admin_binary": "oadm", "all_hostnames": [ "35.184.199.63", "qe-cl-auto-dgwp-master-etcd-zone2-1", "10.240.0.8", "qe-cl-auto-dgwp-master-etcd-zone2-1.0706-5vc.qe.rhcloud.com" ], "cli_image": "openshift3/ose", "client_binary": "oc", "cluster_id": "default", "config_base": "/etc/origin", "data_dir": "/var/lib/origin", "debug_level": "2", "deployer_image": "openshift3/ose-deployer", "deployment_subtype": "basic", "deployment_type": "openshift-enterprise", "dns_domain": "cluster.local", "etcd_runtime": "host", "examples_content_version": "v3.6", "generate_no_proxy_hosts": true, "hostname": "qe-cl-auto-dgwp-master-etcd-zone2-1", "install_examples": true, "installed_variant_rpms": [], "internal_hostnames": [ "qe-cl-auto-dgwp-master-etcd-zone2-1", "10.240.0.8" ], "ip": "10.240.0.8", "is_atomic": false, "is_containerized": false, "is_etcd_system_container": false, "is_master_system_container": false, "is_node_system_container": false, "is_openvswitch_system_container": false, "kube_svc_ip": "172.31.0.1", "no_proxy": ".cluster.local,.svc,qe-cl-auto-dgwp-master-etcd-zone2-1", <--snip-->
Under the assumption that this actually causes problems installing we'll treat this as a blocker.
Tested with openshift-ansible-3.6.140-1.git.0.4a02427.el7.noarch.rpm Master sysconfig still had the no proxy settings: #cat inventory_hosts <--snip--> osm_cluster_network_cidr=11.0.0.0/16 osm_host_subnet_length=8 openshift_master_portal_net=172.31.0.0/16 <--snip--> [root@qe-cl-auto-jyej-master-etcd-zone2-1 ~]# egrep '^HTTP_PROXY|^HTTPS_PROXY|^NO_PROXY' /etc/sysconfig/docker /etc/sysconfig/atomic-openshift-master /etc/sysconfig/atomic-openshift-master:NO_PROXY=,172.31.0.0/16,11.0.0.0/16 [root@qe-cl-auto-jyej-master-etcd-zone2-1 ~]# egrep '^HTTP_PROXY|^HTTPS_PROXY|^NO_PROXY' /etc/sysconfig/docker /etc/sysconfig/atomic-openshift-master-api /etc/sysconfig/atomic-openshift-master-api:NO_PROXY=,172.31.0.0/16,11.0.0.0/16 [root@qe-cl-auto-jyej-master-etcd-zone2-1 ~]# egrep '^HTTP_PROXY|^HTTPS_PROXY|^NO_PROXY' /etc/sysconfig/docker /etc/sysconfig/atomic-openshift-master-controllers /etc/sysconfig/atomic-openshift-master-controllers:NO_PROXY=,172.31.0.0/16,11.0.0.0/16
Working on this again in https://github.com/openshift/openshift-ansible/pull/4762
https://github.com/openshift/openshift-ansible/pull/4794
More changes merged.
Verified with openshift-ansible-3.6.165-1.git.0.fd7125e.el7.noarch.rpm
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHEA-2017:1716