06-17-2020 01:23 PM
I had CML running fine. I had configured 2 interfaces on the CML VM, one bridged and one NAT. The address that got assigned to the CML was from the NAT interface. I turned off this interface and rebooted so the CML would get it's address from the bridged network. This all worked, the CML UI is at 192.168.10.80 rather than at 172.16.192.128. However, when I try to connect to this address I get the login screen, but when I login I get a 502 Bad Gateway error. If I connect to 192.168.10.80:9090 I have no problems. How do I get rid of the 502 Bad Gateway error?
Paolo
06-17-2020 02:06 PM
Look at the changing the IP address and limitation :
06-17-2020 04:32 PM
06-18-2020 05:31 AM
Hi Paolo,
Just checking that you have not tried to update the the system using cockpit, as this is how I messed up my setup and had to create the VM from scratch and reach out to Sales to get my activation cleared.
If you have managed to crash the CML if other way, have a look that you have all services running in cockpit.
06-18-2020 07:19 PM
06-18-2020 02:37 PM
Hi @pgaltieri ,
do you have the virl2-controller.service started (Cockpit > Services > System Service)? Has this happened after the upgrade?
I have the same issue and I have the virl2-controller.service that is getting stopped with "virl2-controller.service: Failed with result 'signal'."
Thank you
06-18-2020 08:52 PM
07-31-2020 08:32 AM
One way to eliminate the problem is to do the following on the CML VM
sudo systemctl restart virl2.target
This eliminated the 502 error for me and allowed me to de-register the license and re-deploy.
Paolo
06-24-2020 05:41 PM
Stefano,
I believe I have found a fix to the 502 bad gateway problem. When it occurs login to the CML OVA and issue the following:
sudo systemctl restart virl2.target
This fixed the problem for me on 2 systems. Including one where I had sent Karlo a request to fix it. I have sent him email with the same information.
Paolo
06-18-2020 09:03 PM - edited 06-18-2020 09:24 PM
Hi, same problem here.
After updating CentOS via the CML Admin-GUI the CML service wouldn't start anymore.
In total there are three services that don't run porperly anymore:
virl2-controller.service
systemd-timedated.service
timedatex.service
With the last two only one will run while the other complains that there is already a service running with the same BusName.
From what I can tell, the virl2-controller.service requires both to be up to run and it is virl2-controller.service that runs the socket which nginx is trying to open (and throwing a 502 bad gateway).
It's too bad the licensing is stuck in the now inaccessible Lab-GUI rather than in the Admin-GUI, this way we can't just set up a new VM.
Hope someone finds a solution to this.
Cheers,
Stephan
06-19-2020 12:37 AM
Hi,
there are multiple threads going on for this issue and there's still no "official" solution.
I had to re-deploy my CML as the update broke the system. I have put some more details here on my conclusions: https://learningnetwork.cisco.com/s/question/0D53i00000XSv2RCAT/virl2controllerservice-failed-with-result-signal-what-is-the-fix-for-this-i-tried-restarting-the-service-in-the-sysadmin-account
At the moment the only solution seems to be to re-deploy and have the license release by the Store Team emailing Karlo (kabobile@cisco.com) please see the Out of Compliance instructions here: https://learningnetwork.cisco.com/s/question/0D53i00000VCdNGCA1/out-of-compliance-issuefix
On the new instance do not upgrade unless instructed to do so on the Cisco Documentation.
Thank you
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide