Bug 1623872

Summary: Independent mode of OCS installation for application with registry tries to create heketi pod once again
Product: [Red Hat Storage] Red Hat Gluster Storage Reporter: SATHEESARAN <sasundar>
Component: CNS-deploymentAssignee: Michael Adam <madam>
Status: CLOSED NOTABUG QA Contact: Prasanth <pprakash>
Severity: high Docs Contact:
Priority: unspecified    
Version: cns-3.10CC: akhakhar, asambast, hchiramm, jarrpa, kramdoss, madam, pprakash, rhs-bugs, rtalur, sankarshan
Target Milestone: ---Keywords: ZStream
Target Release: ---   
Hardware: x86_64   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2019-04-16 12:34:29 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 file used to deploy oc cluster and heketi
none
inventory file used to deploy registry,logging and metrics
none
ansible log file corresponding to oc cluster deploy
none
ansible log file corresponding to oc cluster deploy
none
ansible log file corresponding to registry,logging and metrics none

Description SATHEESARAN 2018-08-30 11:11:19 UTC
Description of problem:
-----------------------
This installation issue is seen with independent mode of installation of OCS 3.10 in a brown field fashion

OC cluster setup was up, post that trying to setup registry, logging & metrics. While doing that used the same playbook ( used to create OC cluster ) with parameters required for registry, logging & metrics.

On the deployment ( playbooks/openshift-gluster/config.yml ),  ansible tries to create heketi pod once again and fails as the bricks are already carved out of /dev/sdd.

Version-Release number of selected component (if applicable):
--------------------------------------------------------------
OCP 3.10
RHGS 3.4.0 nightly ( glusterfs-3.12.2-18.el7rhgs )

How reproducible:
-----------------
consistent ( 6/6 )

Steps to Reproduce:
--------------------
1. Run the ansible playbook - playbooks/deploy_cluster.yml - to deploy cluster ( with the inventory file attached, referred as inventory1 )
2. Add the registry, logging & parameters related parameters to the same inventory file ( inventory2 )
3. Run the ansible playbook - playbooks/openshift-glusterfs/config.yml - to deploy registry, logging & metric related pods.

Actual results:
---------------
Deployment of registry, logging & metrics fails as the ansible once again tries to create heketi pod, which is already available

Expected results:
-----------------
Deployment of registry, logging & metrics should happen successfully



Additional info:

Comment 2 SATHEESARAN 2018-08-30 11:31:24 UTC
Created attachment 1479753 [details]
inventory file used to deploy oc cluster and heketi

Comment 3 SATHEESARAN 2018-08-30 11:32:00 UTC
Created attachment 1479754 [details]
inventory file used to deploy registry,logging and metrics

Comment 4 SATHEESARAN 2018-08-30 11:35:18 UTC
Created attachment 1479755 [details]
ansible log file corresponding to oc cluster deploy

Comment 5 SATHEESARAN 2018-08-30 11:38:25 UTC
Created attachment 1479756 [details]
ansible log file corresponding to oc cluster deploy

Comment 6 SATHEESARAN 2018-08-30 11:47:36 UTC
Created attachment 1479758 [details]
ansible log file corresponding to registry,logging and metrics

Comment 7 Jose A. Rivera 2019-04-16 12:34:29 UTC
You cannot install OCS more than once. By having the [glusterfs] group in the second inventory file you are indicating to the playbook that these are new nodes to be installed.

Closing this as NOTABUG.