Bug 1980302

Summary: VNC console stays in Connecting state.
Product: OpenShift Container Platform Reporter: OpenShift BugZilla Robot <openshift-bugzilla-robot>
Component: Console Kubevirt PluginAssignee: Gilad Lekner <glekner>
Status: CLOSED ERRATA QA Contact: Guohua Ouyang <gouyang>
Severity: urgent Docs Contact:
Priority: high    
Version: 4.8CC: aos-bugs, cnv-qe-bugs, fdeutsch, glekner, gouyang, klistenisjulio, sgott, tnisan, williamsdanielusa, yadu, ycui, yzamir
Target Milestone: ---   
Target Release: 4.8.z   
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: 2021-08-10 11:27:39 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Bug Depends On: 1977037    
Bug Blocks:    

Description OpenShift BugZilla Robot 2021-07-08 10:12:14 UTC
+++ This bug was initially created as a clone of Bug #1977037 +++

Created attachment 1795563 [details]
screenshot of Openshift Console

Description of problem:
Cannot connect to VNC console using the Openshift web console and virtctl command

[21-06-28|15:22:37]~>virtctl vnc rhel8-homeless-rabbit
{"component":"","level":"info","msg":"connection timeout: 1m0s","pos":"vnc.go:144","timestamp":"2021-06-28T15:22:38.122078Z"}
Error encountered: exec: "remote-viewer": executable file not found in $PATH



Version-Release number of selected component (if applicable):
Openshift 4.8.0-rc.1
CNV: 2.6.5
Fresh installation deployed with assisted bare metal ( running on baremetal single node deployment (SNO))

How reproducible:


Steps to Reproduce:
1. Install Openshift 4.8.0-rc.1 ( I have not tested another method than Assisted SNO)
2. Install Openshift CNV operator, create HCO
3. Spin a VM 
In my case I imported a RHEL 8.3 ISO as a boot CDROM. 

Actual results:
Import succeed 
Pod virtlauncher is in Running state. 
VM is in running state
Cannot connect to VNC, web UI shows Connecting 

Expected results:
VNC console shows VM display. 

Additional info:

--- Additional comment from sgott on 2021-06-28 21:09:33 UTC ---

Hi Martin,

Could you attach the yaml file for your VM definition? Also, must-gather output would be very helpful to help us diagnose what happened.

--- Additional comment from ipinto on 2021-06-29 06:35:38 UTC ---

Hi Martin,

CNV 2.6.5 should be installed on OCP 4.7. What did you install with OCP 4.8 RC-0?

--- Additional comment from mouimet on 2021-06-29 13:22:22 UTC ---

Hi, I simply took the version that was in the Operator Hub of the newly installed OCP 4.8. which is 2.6.5... can this issue be related to a mismatch of versions ? Thanks,

--- Additional comment from mouimet on 2021-06-29 13:24:43 UTC ---

Created attachment 1795801 [details]
VM definition

--- Additional comment from mouimet on 2021-06-29 13:26:44 UTC ---

The VM (see attachment) was created using the VM creating wizard.

--- Additional comment from sgott on 2021-06-30 12:14:00 UTC ---

Martin,

Could you please confirm that the VM is actually running. The yaml you shared has "Running = false"

Please install virt-viewer in order to confirm behavior of "virtctl vnc"

--- Additional comment from jlejosne on 2021-07-01 15:21:44 UTC ---

I just installed the latest OCP 4.8 nightly + CNV 2.6.5 + HCO to reproduce this issue and for some reason, in the web UI, "Virtual Machines" does not appear under "Workloads".
I double checked that everything is running fine, I'm not sure what's causing this.
Any idea would be appreciated!
Thanks

--- Additional comment from sgott on 2021-07-01 20:42:45 UTC ---

Yaacov,

Can you help out Jed? per Comment #7.

--- Additional comment from gouyang on 2021-07-02 12:17:49 UTC ---

(In reply to Jed Lejosne from comment #7)
> I just installed the latest OCP 4.8 nightly + CNV 2.6.5 + HCO to reproduce
> this issue and for some reason, in the web UI, "Virtual Machines" does not
> appear under "Workloads".
> I double checked that everything is running fine, I'm not sure what's
> causing this.
> Any idea would be appreciated!
> Thanks

If it's OCP 4.8 nightly, it should has the fixes that could show virtualization on CNV 2.6.5, it's confirmed that OCP 4.8.0-rc.1 works with CNV 2.6.

ref link: https://bugzilla.redhat.com/show_bug.cgi?id=1977355#c2

--- Additional comment from gouyang on 2021-07-02 12:42:07 UTC ---

Confirmed the issue is reproduced on OCP-4.8.0-rc.1 + 2.6.5.

--- Additional comment from jlejosne on 2021-07-02 17:27:32 UTC ---

Found rc1 and finally got a repro.
I confirmed that `virtctl vnc` works fine, so the issue must be UI-related...
I also tried creating a VM in the default namespace, in case the namespace was somehow omitted, but no luck there either.

--- Additional comment from gouyang on 2021-07-07 02:23:43 UTC ---

clear the needinfo.

Comment 5 Guohua Ouyang 2021-08-02 14:10:27 UTC
Reproduced the bug on CNV 2.6.6 + OCP 4.8 and verified on CNV 2.6.6 + OCP upstream/release-4.8.

Comment 7 errata-xmlrpc 2021-08-10 11:27:39 UTC
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 (Moderate: OpenShift Container Platform 4.8.4 bug fix and security update), 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/RHSA-2021:2983

Comment 8 Daniel Williams 2022-11-01 01:59:10 UTC
The fix you shared is very nice and easy to follow, thank you for sharing. https://driftboss.net