Initial console shows no config but xsconsole does...

[David Nock1709157254]

Hi,

 

We have an issue where the initial console screen that is shown on startup shows no configuration details. i.e. no machine name/os version/no network/no virtual machines. However, starting a local shell and running xsconsole shows everything correctly.

 

Is there any way to get this back in sync?

 

All VMs are running and we can manage the server via XenCenter.

 

Currently have XenServer 6.5 installed.

 

Thanks.

[Fabio Papa]

My temporary server running XS7.0 does that too, but it eventually refreshes. Is maybe refreshing disabled? What if you try refreshing it by hand? I think there's a key for that...

[David Nock1709157254]

Have just checked server again and still not showing details. I see on the console there is an <F5> refresh option but that doesn't change anything.

 

The hardware is a HP ProLiant BL465c blade. Other blades in the server show correct details for their VMs etc...

[Boby John1709155536]

HI David,

 

Are you running a custom hardware ?

Also do you have all the patches /firmware up to date ? 

[David Nock1709157254]

This is just a standard blade with up to date firmware.

 

We have the same version of XenServer installed on a number of identical blades and all running fine.

 

I guess ultimately exporting the VMs and running a fresh install/import would get things running but I'm trying to avoid that option.

[Boby John1709155536]

the moment when you are seeing this , Can you connect via ssh and look at the log files and see if you are able to grab something

[David Nock1709157254]

I have ssh access and can browse through the logs, but I don't really know what I am looking for...

[Boby John1709155536]

Look through

 

/var/log/messages

/var/log/xensource.log

/var/log/kern.log

 

and check if you see any error

[Jonathan Johnson]

I have this same problem with XenServer 7.0. I have two SuperMicro servers and one Dell server in a pool; the Dell server works fine but the two SuperMicro servers (different models) exhibit the same behavior.

 

I believe that the XSConsole on the the physical console (tty1) is launching before the toolstack or some other dependency has fully started.

 

As a workaround:

1. Go to a command prompt (menu option, Alt+F3, XenCenter console, or SSH)
2. Identify the XSConsole process:
   

   ps aux | grep -i xsconsole

    

3. This should return a process for a Python script, XSConsole.py running on tty1. Note the process ID, which is the value in the second column.
4. Kill the Python script:
   

   kill <PID>

    

5. The XSConsole on the physical console will automatically respawn, showing correct information and providing full functionality.
6. To return to the XSConsole on tty1 (if you pressed Alt+F3), press Alt+F1.

This should work for 6.x series Xenserver as well.

[Jonathan Johnson]

I have this same problem with XenServer 7.0. I think that what's happening is that xsconsole on the physical console (tty1) is launching before the toolstack or another dependency is fully loaded.

 

As a workaround:

1. Go to a command prompt (menu option, Alt+F3, XenCenter, or SSH)
2. Identity the process ID of the Python script that's running xsconsole on tty1 (ps aux | grep -i xsconsole)
3. Kill the process (kill <PID>)

This will respawn xsconsole on tty1 (the physical console) with correct information and full functionality.

 

(If you pressed Alt+F3 to get a command prompt, press Alt+F1 to return to tty1/xsconsole.)

[Phil S]

Hi!

 

Could you solve that issue? One of my servers is showing the same problem. Restarting xsconsole seems to be a workaround, but I would prefer to solve the issue...

 

Regards,

Phil

[Phil S]

Hi!

 

As a workaround, I added the following code to rc.local:

 

/usr/bin/sleep 30 && /usr/bin/ps -ef | grep xsconsole | /usr/bin/grep -v grep | /usr/bin/awk '{print $2}' | /usr/bin/xargs kill &

 

Regards

[Leonid Mayzenberg]

Just to refresh
3 servers with  CH(XS) 8.1 (NUC based) + CIFS share ( windows based) for HA pool. After power outage CIFS is working well and it's network accessible, all files are in place. But all 3 XS after booting don't see own IPs and can't come back to pool. I tried to reset network connector through Management console  but without any success.
It’s looks like my CIFS based on Windows2012R2 server  Share disk has some network issues , very rarely or during reboot

So, idea was to clear or unplug CIFS disks
Repair steps are below

For Master
xe host-list – not work because IP is absent.
xe pool-emergency-transition-to-master
xe host-is-in-emergency-mode – answered - YES
xe host-emergency-ha-disable --force
xe sr-list name-label=CIFS
xe pbd-list sr-uuid=52fb2fd7-573d-6bf8-c60e-d54b15c5cdbb
See list of all disks on my CIFS
Unplug them
xe pbd-unplug uuid=5b3b5f99-8d67-40c9-3247-f2e8b84fe592
xe pbd-unplug uuid=21f8000d-853d-5c09-17d6-dc07a326965b
pbd-unplug uuid=25ad3ce0-4031-50cb-87e0-9e152788a26b
Check again
xe pbd-list sr-uuid=52fb2fd7-573d-6bf8-c60e-d54b15c5cdbb
All CIFS disk are unplugged
xe-toolstack-restart
xsconsole
Master came back

2 and 3 servers after  xe-reset-networking  and reboot
came back to pool with DHCP addresses. Easy to repair