Bug 1473660

Summary: installer is also trying to install other pure new node when running master scale up playbook.
Product: OpenShift Container Platform Reporter: Johnny Liu <jialiu>
Component: InstallerAssignee: Scott Dodson <sdodson>
Status: CLOSED WONTFIX QA Contact: Johnny Liu <jialiu>
Severity: medium Docs Contact:
Priority: high    
Version: 3.6.0CC: aos-bugs, jokerman, mmccomas
Target Milestone: ---   
Target Release: ---   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2017-08-14 15:51:48 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:
Attachments:
Description Flags
inventory host file and log none

Description Johnny Liu 2017-07-21 11:58:44 UTC 
Description of problem:
installer is also trying to install other pure new node when running master scale up playbook. 

Version-Release number of the following components:
openshift-ansible-3.6.153-1.git.0.5a6bf7d.el7.noarch

How reproducible:
Always

Steps to Reproduce:
1. setup up a master HA env.
2. launch 2 new hosts, one is for new master, another is for new node.
3. on the two instance, subscribe channel, and install excluder.
4. add new_masters and new_nodes into inventory host file
5. run "playbooks/byo/openshift-master/scaleup.yml" to do master scale up

Actual results:
playbook failed.
TASK [openshift_common : Install the base package for versioning] **************
Friday 21 July 2017  11:03:44 +0000 (0:00:01.575)       0:07:29.630 *********** 

ok: [qe-jialiu-master-new-1.0721-04f.qe.rhcloud.com] => {
    "changed": false, 
    "rc": 0, 
    "results": [
        "atomic-openshift-3.6.153-1.git.0.4894417.el7.x86_64 providing atomic-openshift-3.6.153 is already installed"
    ]
}
fatal: [qe-jialiu-node-new-1.0721-04f.qe.rhcloud.com]: FAILED! => {
    "changed": false, 
    "failed": true, 
    "rc": 126, 
    "results": [
        "No package matching 'atomic-openshift-3.6.153' found available, installed or updated"
    ]
}

MSG:

No package matching 'atomic-openshift-3.6.153' found available, installed or updated

That is because master scaleup also involve other pure new_nodes, while openshift_excluder does not do unexclude operation.

openshift_excluder role is doing right things, only do unexlude operation on master. while installer is also trying to install other pure new nodes.

Expected results:
when running master scaleup, installer should only install master and node service this new_masters, should NOT install other pure new node.
pure new node install work should be left to playbooks/byo/openshift-master/scaleup.yml.


Additional info:
Comment 1 Johnny Liu 2017-07-21 12:11:22 UTC 
Created attachment 1302331 [details]
inventory host file and log
Comment 2 Scott Dodson 2017-08-14 15:51:48 UTC 
I'm going to mark this as CLOSED WONTFIX. When performing scaleup operations it's expected that only new_nodes or new_masters exist in the inventory and if that were the case this problem wouldn't happen.