OES2 Linux Activation Fails or OES2 Linux says there are no updates available

Another issue related to the cloning of virtual servers from a VMWare template. Don't get me wrong, I love VMWare, but the human element involved in cloning servers (especially after some time has passed) always shoots me in the foot.

I stumbled upon this when I was doing updates for a customer. A few servers updated just as they should, while others were saying there were no updates available.  Then when re-activating them via Novell Customer Center I realized I'd used the wrong server name. Doh! SMH - again. One of those steps I overlooked after cloning the server. Part of this depends on your particular template process, but if you're not controlling the process this "fix" is a good thing to have in your toolbelt.

This comes from a Novell TID that was not easy to find. Hopefully this post will make generic searches outside of the Novell KB a bit easier.

These are the error messages you may encounter:

• ERROR: Could not add'https://update.novell.com': Failed to download XML metadata: Download failed: (https://update.novell.com/repo/repoindex.xml) serializationStream supports seeking, but its length is 0
  
  ERROR: Could not add 'nu.novell.com': Failed to parse XML metadata: Can't add repository at https://nu.novell.com/repo/$RCE/SLED10-Updates/sled-10-i586: Unknown Error: Downloading metadata failed (is YUM source?) or user did not accept remote source. Aborting refresh.
  
  ERROR: Could not add 'nu.novell.com': Failed to parse XML metadata: Can't add repository at https://nu.novell.com/repo/$RCE/SLES10-Updates/sles-10-i586: Unknown Error: Downloading metadata failed (is YUM source?) or user did not accept remote source. Aborting refresh.

My trigger was that the servers were saying no updates were available and it had been 6 months since I'd touched them and they weren't setup for automatic updates.  This is apparently due to a corrupt cache.

Here's how to fix it:

1. Close the Novell Customer Center Configuration tool if it's running
2. Open a terminal window and su if you're not already logged in as root
3. Run the following commands
• rczmd stop
• rm -rf /var/cache/zmd/*
• rczmd start
4. Launch Yast2  and select Software | Novell Customer Center Configuration
5. Correct the server name if it's wrong and finish the activation process

This is all laid out in Novell TID #3818394 with the title of "Activation fails with an error downloading XML data" if you want to see the official resolution.

Cannot create NSS volume on OES2 Linux box

This was one of the aggravating issues I ran into over this past Summer/Fall.  It was my own fault, but trying to find a solution was not easy.

Here's the assorted searches I performed:

• cannot install an NSS volume to eDirectory
• ERROR: NSS Admin Failed Creating Edirectory Objects
• Error -676 When Trying To Create A Pool Or Volume via NSSMU or iManager
• cannot create NSS volume on OES2 Linux on VMWare

This happens more often than I care to admit on virtual servers.

If you have cloned the server in question from another server - for instance using VMWare templates - you may be using the same NSS admin name as the orginally created server. This is the real issue. It's not a VMWare issue and it's not really an OES2 Linux issue, it's a doh!, slap my forehead issue.

Here's how to fix it:

1. Open up a terminal window
2. Type the command "id novlxregd" and confirm that the nssadmin name is wrong
3. Go into Yast | OES Install and Configuation. 
4. "Enable" reconfiguration under the NSS section. 
5. If the name doesn't match your server name, i.e. cn=server2admin,o=novell if the server with the issue is named server2, simply run through the NSS configuration after changing the name to the corrected name
6. Let it finish and the type the "id novlxregd" again and you should see the correct response now.

I ran into this after having used a template for a multiple server migration for a customer. First one that I'd missed this on, but it wasn't my last.