Forum Discussion

Anthony's avatar
Anthony
Icon for Nimbostratus rankNimbostratus
Oct 19, 2012

upgrading in stages from 10.2.0 - loss of node names

Hi all, I appologise if this is posted in the wrong area, but it is config based.

 

 

I have the task of upgrading our 2 test & 2 production loadbalancers.

 

 

All are currently on LTM v10.2.0.

 

 

I have taken the approach of moving up in steps of main versions and final hotfixes as my previous attempt to go to 11.2.0 failed.

 

 

When I upgraded to 11.0.0, I can see that the way nodes are handled has changed. As such, everything showing in our network map is no longer itendifyable in the usual manor. We have a large amount of nodes so being able to view them by a meaningful name is important.

 

 

I found one article that suggested editing the bigip.conf, but having done that I get an error when I try to load it again - so I'm wondering if I'm missing something there.

 

 

This is the error on loading the bigip.conf...

 

 

[root@romulus:Standby] config tmsh load sys config

 

Loading system configuration...

 

/defaults/app_template_base.conf

 

/defaults/config_base.conf

 

/config/low_profile_base.conf

 

/defaults/wam_base.conf

 

/usr/share/monitors/base_monitors.conf

 

/config/daemon.conf

 

/config/profile_base.conf

 

/defaults/fullarmor_gpo_base.conf

 

Loading configuration...

 

/config/bigip_base.conf

 

/config/bigip_user.conf

 

/config/bigip.conf

 

01070734:3: Configuration error: invalid node address, the IP address 192.168.10.11 already exists

 

Unexpected Error: Loading configuration process failed.

 

 

Can anyone advise me if there are any other ways of getting out node names back? Or give more information on how I can update them in the bigip.conf if it is a manual task?

 

 

Many thanks,

 

Anthony

 

5 Replies

  • 01070734:3: Configuration error: invalid node address, the IP address 192.168.10.11 already existsis there duplicated 192.168.10.11 in bigip.conf?
  • In the node list, no. The address is listed multiple times in the pool lists though as we use it with verying ports for different services.

     

     

    Thanks

     

    Ant

     

  • if you think config file is correct already (no duplicated node), can you try this?

     

     

    sol13030: Forcing the mcpd process to reload configuration

     

    http://support.f5.com/kb/en-us/solutions/public/13000/000/sol13030.html
  • What you might also want to try;

     

     

    1) Reorder the node configuration statement so they appear before or after the pool members (whatever the opposite of now is)

     

    2) Remove the node configuration from the file, load the configuration and then paste the node configuration in separately

     

    3) The opposite, paste the node configuration first, remove from the import file, then load the file
  • Thanks all for you input. I have taken the approach of adding a new VS, Pool & Node and then having a look at how the config is constructed for these new items.

     

     

    From that I can see that I only need to replace the IP address on the Node, and on the /Common/: of the Pool. I've just made 1 simple change and reloaded the config and the change is there! I can now work my way through each of them to update the file in full.

     

     

    23 nodes and 58 pools... think I konw what I'll be doing till the end of the day!

     

     

    Thanks again for your inputs.

     

     

    Ant