VM machines have strange status' since update.

Submitted by mmi on Fri, 03/06/2015 - 01:36 Pro Licensee

Since we updated from 7.9 to 8.0 Almost all hosts have several status' which make no sense at all.

SSH not installed. Down (Host) Webmin not installed Ping Failed

And mind you this is across several different physical hosts. A reboot sometimes solves the problem but if it does 5 minutes later the stats are back.

Are there any known issues in regard to this?

Thank you in advance.

Status: 
Active

Comments

Submitted by mmi on Fri, 03/06/2015 - 09:25 Pro Licensee

Extra information:

Also Status No SSH is there a lot, which makes also no sense at all. If we do an ssh to a server it has connection within a second. also root password is ok since they were distributed by cloudmin itself.

Also with a lot of the servers the message Could not fetch Citrix Xen VMs : Timeout connecting appears. which is also new to us. there are no issues with ping/ssh to the xenservers. Even the names are in the /etc/hosts to prevent connection loss when dns fails.

Submitted by JamieCameron on Fri, 03/06/2015 - 10:44

Having multiple statuses is very odd - did this start happening when you upgraded Cloudmin from 7.9 to 8.0, or was it an upgrade on the VMs that triggered it?

Submitted by mmi on Wed, 03/11/2015 - 02:59 Pro Licensee

Yes we have two environment developers and production; Development cloudmin was in a ealier 7.9. The production cloudmin is from the beginning on 8. The production had this strange issues already from the beginning. After updating the developemnet cloudmin this strange behavior started there also. Maybe it's coinsidence which we can only test if we roll back to 7.9. Can you provide maybe a way to do this without loosing the configuration/passwords etc. etc.. We could also install a new Cloudmin from scratch with 7.9 but then it would be nice we can copy the configuration to that server. At least if we try that we can maybe exclude if it is a module problem. Thanks in advance

Submitted by JamieCameron on Wed, 03/11/2015 - 20:18

You can safely roll back to 7.9 with no loss of settings - just SSH in as root and run yum install wbm-server-manager-7.9

Submitted by mmi on Thu, 03/12/2015 - 11:05 Pro Licensee

Hi

If I run the command it returns package matching 2:wbm-server-manager-7.9.1.noarch already installed.......nothing to do The cloudmin interface tells me that I am running 8.0 Description Webmin module for 'Cloudmin' in RPM format Package wbm-server-manager Class System/Tools Version 8.0-1 Vendor Jamie Cameron Architecture noarch Installed 26/Feb/2015 17:02 other ways to roll back?

Submitted by JamieCameron on Thu, 03/12/2015 - 13:17

If you run rpm -q wbm-server-manager , which version does it shows is installed?

Submitted by mmi on Thu, 03/12/2015 - 16:16 Pro Licensee

hi The result is: wbm-server-manager-8.0-1.noarch

Submitted by JamieCameron on Thu, 03/12/2015 - 18:59

Try instead running yum downgrade wbm-server-manager-7.9

Submitted by mmi on Fri, 03/13/2015 - 09:26 Pro Licensee

thanks that works, i will report the results. mario

Submitted by mmi on Fri, 03/13/2015 - 09:42 Pro Licensee

Hmm that seems to work for the last 10 minutes, no strange behavior in development like 'no-ssh' etc. Do you have a suggestion regarding downgrading cloudmin on ubuntu 12.04 prod cloudmin server? thanks in advance

Submitted by JamieCameron on Fri, 03/13/2015 - 15:31

Ok, so clearly something has changed between 7.9 and 8.0 which is causing the status to be incorrectly collected.

I would like to track down the underlying issue here - If you have a system that is still seeing this, is there any chance I could login and see what is going wrong internally?

Submitted by mmi on Wed, 03/18/2015 - 08:43 Pro Licensee

Hi I can give you access via fortinet vpn/ssh via fixed ip to the dev environment server with the downgraded cloudmin which is working right now.You could upgrade it to version 8 again and see what happens. Or we could clone it for the moment and upgrade one cloudmin server to 8. (please tell me what you prefer) PS.After the downgrade the status is stays perfectly The production environment however is another story, since we did not enrolled the certificates yet, so the management server is not able to have a good status of the VM's at all (just ping ssh login failed etc...). So giving access to that does not make sence.

PS. Can you check comment #10

thanks

Submitted by JamieCameron on Wed, 03/18/2015 - 14:55

Sure, remote access to a system (even a test environment) that is showing this problem would be really useful. Email me at jcameron@virtualmin.com with the login details.