in

myITforum.com

Brian Hofmeister at myITforum.com

Doing less with less...
  • Cisco ASA... Dead

    I finally got some time to investigate one of our Cisco Adaptive Security Appliances that was returned from a plant.  I've never seen this problem before with an ASA so I thought I would post my findings in case anyone else has this problem.

    I hopped on the console and found out quickly it was stuck in a reboot cycle.  Here was the output:

    Booting system, please wait...

    CISCO SYSTEMS
    Embedded BIOS Version 1.0(10)0 03/25/05 22:42:05.25

    Evaluating BIOS Options ...
    Launch BIOS Extension to setup ROMMON

    Cisco Systems ROMMON Version (1.0(10)0) #0: Fri Mar 25 23:02:10 PST 2005

    Platform ASA5510

    Use BREAK or ESC to interrupt boot.
    Use SPACE to begin boot immediately.
    Launching BootLoader...
    Boot configuration file contains 1 entry.

    Loading disk0:/asa722-k8.bin... Booting...
    256MB RAM

    Total SSMs found: 0
    nic_plat_add_io_cards(): Unable to get the SSM-4GE slot info

    Total NICs found: 5
    mcwa i82557 Ethernet at irq 11  MAC: 0011.e939.019a
    mcwa i82557 Ethernet at irq  5  MAC: 0000.0001.0001

    Booting system, please wait...

    I put in a case with TAC and they told me I probably had a bad boot image.  I booted to ROMMON, tried to run the commands for setting up a TFTP transfer, and found myself hitting another wall.  Here are a few of the things I tried:

    rommon #5> dev
    Interface Device Information:
      Ethernet0/0: i82546GB, PCI: bus-3, slot-3, fct-1, rev-3, irq-9
      Ethernet0/1: i82546GB, PCI: bus-3, slot-3, fct-0, rev-3, irq-9
      Management0/0: i82551ER, PCI: bus-4, slot-2, fct-0, rev-16, irq-11

    rommon #0> PORT=Management0/0
    Backplane Data Channel
    Link is DOWN
    MAC Address: 0000.0001.0002

    rommon #8> interface Ethernet0/1
    ef_reset: i82546GB_copper Init failed
    Ethernet0/1
    Link is DOWN
    MAC Address: 0011.e939.019a

    rommon #0> interface Ethernet0/0
    (*This caused the ASA to freeze, had to reboot)

    Everything was showing "LINK DOWN". I verified I had a good cable/network drop by plugging in my laptop to the same cable/network drop.  For good measure, I tried a different cable/network drop as well.

    My guess is that this thing is a paper weight.  If anyone has any suggestions or sees something I am overlooking, please drop me a note.

    Posted Dec 01 2007, 11:04 AM by Anonymous with no comments
    Filed under:
  • MS Virtual Server - 1, VMWare - 0

    Well the title might be a little misleading but I had to capture your attention somehow. I think the actual score might be something like "MS Virtual Sever - 1, VMWare - 24".

    Anyone that knows me is well aware of my VMWare allegiance but I think they are really dropping the ball on one particular area.  Currently, VMWare is unable to resize vmfs volumes.  This is a HUGE problem for users of the ESX product. 

    Let me give a scenario that I have run in to several times already.

    I carve out some LUN space for my ESX farm and think, "No way I will ever need more than 250GB for this particular farm."  I connect my ESX servers to the lun and they format the lun using vmfs which is set at 250GB.  A few months go by and guess what, I use up 220GB of space and the remaining 30GB is practically unusable since my standard Virtual Machines are 40GB in size.

    VMWare's solution?  Well actually they have two, one of which is NOT recommended by anyone at VMWare.

    1.  Create a new lun and present it to the ESX farm.  Use the nifty EXTENT feature within the ESX farm to basically Band-Aid the new lun to the existing lun and they work as one big vmfs volume.

    Sounds good eh?  NOT RECOMMENDED by anyone at VMWare.  Just like it sounds, you are basically running your volumes in RAID0.  Half of your VM files may be sitting on 1 lun while the other half of the files sit on the other lun.  What happens if one of the luns has a hiccup and goes offline for a second.  It's basically the equivalent of unplugging one disk in a RAID0 array for one second and then plugging it back in.  Do you think the VMs will live?  Most likely not.

    2.  Create a new lun that is bigger than the original lun.  Take all your VMs offline and COPY them to this new lun.  Then delete the old smaller lun.

    Are you serious?  We have this brilliant piece of technology that is highly redundant and available but if I run out of space, I have to take the whole thing offline?

     

    So all you Microsofties have been reading this article and wondering, "Why does this mean MS Virtual Server is "1 up" on VMWare?"

    Microsoft Virtual Server runs its VMs on NTFS.  An NTFS volume can be resized dynamically without taking the servers offline.  I think every storage vendor has a different tool for doing this but ours is SnapDrive.

    One last thing before I finish this post... I don't like how VMWare misleads you on the web site.  Click HERE for details on VMFS.  They say:

    Dynamic increase of VMFS volume size.  Create new virtual machines without relying on a storage administrator.  Adaptive block sizing and addressing for growing files allows to increase a VMFS volume on the fly.

    That's very misleading considering the available tools/options for resizing.  This blurb should have an asterisk that says "if you like RAID0 and not being supported by VMWare".

Copyright - www.myITforum.com, Inc. - 2010 All Rights reserved.
Powered by Community Server (Commercial Edition), by Telligent Systems