Bug 1428532

Summary: 3.5.0.37: Task Ensure etcd datadir exists fails in containerized install due to non-existent etcd user
Product: OpenShift Container Platform Reporter: Walid A. <wabouham>
Component: InstallerAssignee: Scott Dodson <sdodson>
Status: CLOSED CURRENTRELEASE QA Contact: Johnny Liu <jialiu>
Severity: high Docs Contact:
Priority: unspecified    
Version: 3.5.0CC: aos-bugs, jokerman, mifiedle, mmccomas, wabouham
Target Milestone: ---   
Target Release: ---   
Hardware: Unspecified   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: No Doc Update
Doc Text:
undefined
 Story Points: ---
Clone Of: Environment:
Last Closed: 2017-03-08 14:12:45 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:

Description Walid A. 2017-03-02 18:49:03 UTC 
Description of problem:

On AWS EC2, Multi-master OCP HA containerized BYO advanced install fails trying to chown non-existent etcd user.  Error occurs in

roles/etcd/tasks/main.yml task:

TASK [etcd : Ensure etcd datadir exists] ***************************************
Thursday 02 March 2017  17:12:08 +0000 (0:00:00.136)       0:06:10.129 ******** 
fatal: [<Public DNS Name>]: FAILED! => {"changed": false, "failed": true, "gid": 0, "group": "root", "mode": "0755", "msg": "chown failed: failed to look up user etcd", "owner": "root", "path": "/var/lib/etcd/", "secontext": "unconfined_u:object_r:var_lib_t:s0", "size": 6, "state": "directory", "uid": 0}

Version-Release number of selected component (if applicable):

oc v3.5.0.37
kubernetes v1.5.2+43a9be4

This issue may not be specifically related to HA cluster install, but happened twice in a row during HA cluster install.

How reproducible:
Reproduced twice in a row.

Steps to Reproduce:
1. Set openshift-ansible var containerized to true
2. Run openshift-ansible BYO config.yml playbook to install a containerized Multi-Master OCP cluster (3 masters/etcds, 1 load balancer, 2 infra node, and 1 application node).  


Actual results:
Install fails at TASK [etcd : Ensure etcd datadir exists]

Expected results:
Install to complete with no errors

Additional info:
openshift-ansible git commit:
git rev-parse HEAD
5a91f31b65a4bb0ec524aee9ba1c6d4e4030d8d2
Comment 2 Mike Fiedler 2017-03-02 19:57:31 UTC 
Single master containerized installs are failing as well running openshift-ansible latest.
Comment 3 Scott Dodson 2017-03-03 02:57:59 UTC 
https://github.com/openshift/openshift-ansible/pull/3551

should fix this
Comment 5 Walid A. 2017-03-03 17:24:30 UTC 
Verified on OCP 3.5.0.37, with this morning's openshift-ansible git pull (revision b777f84e9e86558e9083b5ce8e62d606f955be55).

HA containerized cluster installed successfully.