Thirty Helens Agree

I had an issue with an iSCSI target LUN not being formatted as VMFS in ESX Server 3.0.X.  The problem turned out to be on my side, but I ended up with a number of blank SCSI Targets under VI -> Configuration -> Storage Adapters.  My SCSI Targets jumped from SCSI Target 2 to SCSI Target 9.  This was not an ideal situation as I want all of the members of the DRS/HA cluster to have the same configuration.  The way to fix up the SCSI Targets involves adjusting some of the configuration files on the ESX Server itself.  First you need to move any and all running VMs off of the host you’re going to be working on and put it into Maintenance Mode.

• Log into the service console as root.
• cd /var/lib/iscsi and there are two control files there: vmkbindings, vmkdiscovery
• temporarily disable iSCSI: esxcfg-swiscsi -d
• kill the iSCSI software: esxcfg-swiscsi -k
• edit both the vmkbindings and vmkdiscovery to remove the LUNs that are no longer valid.  You can change the target ID as well if you want to move that around.
• start iSCSI back up: esxcfg-swscsi -e
• wait about 30 seconds and do a rescan from the command line: esxcfg-swscsi -s
• once you do that you can go into the VI Client and rescan the Storage Adapters.  If this looks right now you’re all set.
• take the server out of maintenance mode and move a VM to the server to make sure it is working correct

That’s it!  Good luck, this saved me a decent amount of time and I wasn’t able to find this information in the forums or on the interwebs.

They asked the new host of The View is she thought the world was flat.  She honestly doesn’t KNOW if the world is flat or not.  Watch her wiggle out of this one.  Unbelievable. How can someone be this stupid?

This has really just been my luck so far this week.  I turn on our new Avocent MergePoint 5300 to start the configuration.  Go into the configuration via the console port:

  1:Summary
  2:Network config
  3:Set date and time
  4:Set admin password
  5:Reboot
  6:Power off
  7:Roll back the version
  8:Security config
  9:Debug Log
  10:Restore to Factory Defaults
  0:Exit
  Please select [0 - 10]:

Fair enough, right?  All I need to do is set up the IP address so I can jump into the configuration website so I select “Network Config”.

  Set eth1 Speed:
  1:Auto-Negotiate
  2:10 Mbps/Half Duplex
  3:10 Mbps/Full Duplex
  4:100 Mbps/Half Duplex
  5:100 Mbps/Full Duplex
  0:Return
  Current Setting [Auto-Negotiate]
  Please select [0 - 5]:1

Set eth1 DHCP?(Y/N):N
Invalid response, please input again [YES/NO] -> NO
Invalid response, please input again [YES/NO] -> N
Invalid response, please input again [YES/NO] -> no
Invalid response, please input again [YES/NO] -> n
Invalid response, please input again [YES/NO] ->
Invalid response, please input again [YES/NO] -> YES
Invalid response, please input again [YES/NO] -> yes
Invalid response, please input again [YES/NO] -> y
Invalid response, please input again [YES/NO] -> quit
Invalid response, please input again [YES/NO] -> exit
Invalid response, please input again [YES/NO] ->
Invalid response, please input again [YES/NO] ->
Invalid response, please input again [YES/NO] ->

Apparently there is no known valid response.  How does a product leave QA in this state?  I already hate this thing and I haven’t even set it up yet. 

This should be more obvious than it is: Launch “Internet Connect”, select “Options…” from the “Connect” menu and deselect “Send all traffic over VPN connection”.

One of my favorite movies (save your jabs for the comments section) Princess Bride is twenty years old. That makes me feel, well, old I guess. I saw this on Digg and the “Then and Now” pictures come via ABC News.