Note: This bug is displayed in read-only format because the product is no longer active in Red Hat Bugzilla.

Bug 1530931

Summary: vphere configuration impacts cluster network
Product: OpenShift Container Platform Reporter: Jaspreet Kaur <jkaur>
Component: NetworkingAssignee: Dan Williams <dcbw>
Status: CLOSED DUPLICATE QA Contact: Meng Bo <bmeng>
Severity: urgent Docs Contact:
Priority: urgent    
Version: 3.7.0CC: aos-bugs, atragler, bbennett, dcbw, dphillip, emahoney, jfoots, jkaur, jokerman, mmagnani, mmccomas, paul.vanallsburg, sdodson
Target Milestone: ---   
Target Release: 3.9.0   
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: 2018-01-31 23:11:51 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
master after vsphere config
none
master before vsphere config
none
router after vsphere config
none
router before vsphere config none

Description Jaspreet Kaur 2018-01-04 08:52:23 UTC
Description of problem: installed Openshift with a basic install script. 
The first time I installed Openshift with routers & registry
then restarted the masters and nodes.  After that the internal registry was fine.

# RegistryAddr=$(oc get svc docker-registry -n default -o 'jsonpath={.spec.clusterIP}:{.spec.ports[0].port}')
[root@vmlxopenas02 ~]# curl -vk $RegistryAddr/healthz
* About to connect() to 10.128.130.5 port 5000 (#0)
*   Trying 10.128.130.5...
* Connected to 10.128.130.5 (10.128.130.5) port 5000 (#0)
> GET /healthz HTTP/1.1
> User-Agent: curl/7.29.0
> Host: 10.128.130.5:5000
> Accept: */*
>
< HTTP/1.1 200 OK
< Cache-Control: no-cache
< Date: Wed, 03 Jan 2018 14:20:24 GMT
< Content-Length: 0
< Content-Type: text/plain; charset=utf-8
<
* Connection #0 to host 10.128.130.5 left intact
[root@vmlxopenas02 ~]#

I rolled back to my VM snapshots and re-installed with a basic install script. 
This time I installed Openshift with routers & registry then added the config for vsphere
then restarted the masters and nodes.  After that the route to the internal registry was lost

# RegistryAddr=$(oc get svc docker-registry -n default -o 'jsonpath={.spec.clusterIP}:{.spec.ports[0].port}')
[root@vmlxopenas02 ~]# curl -vk $RegistryAddr/healthz
* About to connect() to 10.128.154.196 port 5000 (#0)
*   Trying 10.128.154.196...
* No route to host
* Failed connect to 10.128.154.196:5000; No route to host
* Closing connection 0
curl: (7) Failed connect to 10.128.154.196:5000; No route to host

[root@vmlxopenas02 ~]# oc get all -o wide
NAME                                REVISION   DESIRED   CURRENT   TRIGGERED BY
deploymentconfigs/docker-registry   1          3         3         config
deploymentconfigs/router            1          3         3         config

NAME                   DESIRED   CURRENT   READY     AGE       CONTAINER(S)   IMAGE(S)                                 SELECTOR
rc/docker-registry-1   3         3         3         1h        registry       openshift3/ose-docker-registry:v3.7.14   deployment=docker-registry-1,deploymentconfig=docker-registry,docker-registry=default
rc/router-1            3         3         3         1h        router         openshift3/ose-haproxy-router:v3.7.14    deployment=router-1,deploymentconfig=router,router=router

NAME                  CLUSTER-IP       EXTERNAL-IP   PORT(S)                   AGE       SELECTOR
svc/docker-registry   10.128.154.196   <none>        5000/TCP                  1h        docker-registry=default
svc/kubernetes        10.128.0.1       <none>        443/TCP,53/UDP,53/TCP     1h        <none>
svc/router            10.128.21.82     <none>        80/TCP,443/TCP,1936/TCP   1h        router=router


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


How reproducible:


Steps to Reproduce:
1.
2.
3.

Actual results: cluster network impacted after vpshere configuration


Expected results: It should work hgealthy with vsphere configuration


Additional info:

Comment 4 davis phillips 2018-01-04 19:30:52 UTC
When you added the config for vsphere, did you add in the short hostname for the openshift_hostname var in the inventory? 

The cloud provider makes the hostname the VM name and if it does'nt match with the install there can be fallout issues from the CNI networking config.

Comment 10 emahoney 2018-01-08 22:55:30 UTC
Created attachment 1378771 [details]
master after vsphere config

Comment 11 emahoney 2018-01-08 22:55:52 UTC
Created attachment 1378772 [details]
master before vsphere config

Comment 12 emahoney 2018-01-08 22:56:19 UTC
Created attachment 1378773 [details]
router after vsphere config

Comment 13 emahoney 2018-01-08 22:56:39 UTC
Created attachment 1378774 [details]
router before vsphere config

Comment 26 Dan Williams 2018-01-31 23:11:51 UTC
After analysis this bug appears to be a duplicate of https://bugzilla.redhat.com/show_bug.cgi?id=1534720 (or at least the same root cause where a node's IP address changes after it has registered itself).

*** This bug has been marked as a duplicate of bug 1534720 ***