A few months ago I started the process of getting all my ducks in a row for a migration that involved moving all physical servers a customer had to a VMware ESX environment. This environment wasn’t anything special: 60 users running Microsoft Active Directory, Exchange 2003, Blackberry Enterprise Server and Sage Timberline Office as their accounting app.

The Microsoft parts were easy.  I was sure there would be no problem those. My one concern was with the Sage Timberline software package. This program never ran particularly well, and its obvious it is running using very old programming languages much like alot of the LOB apps companies use these days.

I needed to make sure this package would both a) run within a VMware environment and b) be supported by Sage from a technical perspective. When I called the Sage Customer Support desk and asked their response was:

Sage Timberline Office is not supported running as a virtual machine, however we will troubleshoot any issues you have to a reasonable extent, while not allowing any support calls to be escalated to the engineering department.”

That actually sounds pretty fair. The woman I spoke with sounded fairly knowledgable about the Sage product and had been one of the technicians to previously work with another Sage customer that had issues running their software in a virtual environment. I asked her what the specific issues encountered were and all she really could tell me was that the Pervasive Database performed very badly and lots of unexpected things would happen while using the software.

So here I was – ready to get a signature for a proposal that would cost my customer 10’s of thousands of dollars and I had no factual evidence that showed that the accounting package they used would even work. If there were ever going to be a problem with this software it would be related to performance alone, right? There only about 10 users that access this application, so that shouldn’t really be an issue, right?

Well it turns out Sage Timberline Accounting software runs just fine in a virtual environment. It runs better running in VMware ESX 4 than it ever did on its own dedicated Dell Poweredge 1650 with RAID5.

I guess the moral of the story is that if it should work and it isn’t, you’re probably doing something wrong. Fight through the problems with a sense of resolve, do your research, follow guidelines and procedures and most things will run just fine in a virtual environment….and never take no for an answer.

Anyone familiar with ESX will know how finicky it and vCenter can be about name resolution for all parties involved.  I typically make my VI and vCenter Server HOSTS file identical.  This usually solves all my HA problems and other weird issues that arise.

Today I realized it’s also beneficial to duplicate the HOSTS file on a machine that is being converted from physical to virtual using vCenter Converter (the machine I am converting). Before I appended those entries to the I was receiving “FAILED: A system error encountered during an operation.” each time I tried running the conversion.  The error cleared up right away.

I still don’t know why this continues to be an issue. The infrastructures are setup as vanilla and as “by the book” as possible.  DNS and WINS are always setup and working properly.  Short name and FQDN resolution always works from every host and client on the network.

I suppose it isn’t the end of the world. Seems like it should be an easy thing for VMware to “fix”…then again maybe they don’t consider it to be broken.

So I’ve had the problem in the past where I need to remove very old snapshots for ESX virtual machines and eventually after hours of merging it fails with no apparent reason.  Recently I was looking into how I could get aroudn this.  The idea I was testing was to use Converter Enterprise to esentially do a V2V.  That way, if the conversion failed I will still have the original machine in its old working state.

After starting this process on a test VM, after 3 hours of converting it failed.  The logs gave me the following:

UNKNOWN_METHOD_FAULT(vim.fault.NotAuthenticated)

I stumbled upon a post on the vmware communities that suggested I try the following

When setting up the conversion, don’t use “convert all disks and leave current size”, use “select volumes…” and leave all volumes selected and the default options checked.

I gave it another try and voila! Conversion worked without error.

I wonder if multiple disks has anything to do with the snapshot merging problems I’ve had so many times.  As I attempt to remember specifics of all past failed attempts, I believe they’ve all had multiple disks.  Maybe that was the problem all along?  As long as this Plan B works I’m happy!

Strange VMware Tools issue while upgrading one of my Ubuntu 6.06.2 VMware Guests to Ubuntu 8.04.1.  The Ubuntu upgrade itself seemed to go just fine, but installing the updated VMware Tools gave me an error:

Your kernel was built with "gcc" version "4.2.3", while you are
trying to use "/usr/bin/gcc" version "4.2.4". This configuration
is not recommended and VMware Player may crash if you'll continue.
Please try to use exactly same compiler as one used for building
your kernel. Do you want to go with compiler "/usr/bin/gcc"
version "4.2.4" anyway? [no]

It didn’t take very long to find a thread about this – http://ge.ubuntuforums.com/showthread.php?t=963825.  I took the suggest of some of the users here and chose YES.  SNAPSHOT YOUR MACHINES BEFORE YOU TRY SOMETHING UNPROVEN!

Well now it complained that it couldnt locate the Linux C Header files:

The header files in /usr/include are generally for C libraries, not for the
running kernel. If you do not have kernel header files in your /usr/src
directory, you probably do not have the kernel-source package installed. Are yousure that /usr/include contains the header files associated with your running
kernel? [no]

Weird.  So now I had to manually install the Linux Headers.  I did this by issuing:

apt-get install linux-headers-`uname -r` build-essential

I now re-ran perl /usr/src/vmware-tools-distrib/vmware-install.pl.  I again of course had to tell the installer to continue compiling using gcc 4.2.4, but now it didnt complain about the C headers. The install completed successfully.

I had completely forgotten about this. WIth past ESX builds Ubuntu was only *sort of* supported.  In order to get VMWare Tools installed on Ubuntu you had to basically merge the native VMWare Tools with open-vmware-tools.

The process of merging the two sets to tools together was simple enough, and successful enough – but doesnt it feel great when something is really supported and when it works the way it’s supposed to? For the shiny new supported VMWare Tools I simply ran ./vmware-install.pl from the console of my Guest, waited about 5 minutes and voila!

ESX 3.5 Update 3 Release Notes

I was doing some regular maintenance today in Virtual Center 2.5 today and realized Its been a couple months since I’ve installed any updates on my ESX 3.5 hosts.  I don’t think I’ve installed any updates since way back before August before the bugged Update 2 was release. Back then I had reinstalled VI 2.5 and forgot to reinstall Update Manager in VI.  Funny how you forget about things when they aren’t right in front of you.

When I first attempted to Scan for Updates I received the following error in VI:

metadata for patch missing

Searching through the forums suggested the ESX host FW wasn’t properly listening for connections and suggested I run:

esxcfg-firewall --openport 80,tcp,out,dynamicupdates
(port 80 was specific to my environment and may vary for your setup)

Once I did this the Scan for Updates ran just fine

Onto the upgrade….

Before the upgrade

[root@esx-03 root]# vimsh -n -e 'hostsvc/hostsummary' | grep fullName
File not found.
fullName = "VMware ESX Server 3.5.0 build-82663",

1.Reinstalled Update Manager and attached the Critical and Non-critical Baselines to the ESX hosts

2.Put my first ESX host into Maintenance Mode and agreed to all the informational messages that come with that

3.Selected all Guest VMs and migrated them to another ESX host (I’m still not absolutely sure if entering Maintenance Mode actually evacuates all the running VMs.  I just migrate them manually because I can see it working)

4.Right clicked on the ESX host and chose Remediate, chose both Baselines and Next’ed through the wizard.

5.30 minutes later the ESX host was back up and running with the newest patches and updates. Repeat steps on all other Hosts.

After the upgrade

[root@esx-03 root]# vimsh -n -e 'hostsvc/hostsummary' | grep fullName
File not found.
fullName = "VMware ESX Server 3.5.0 build-123630",

One thing I noticed after the upgrade was that the SSH server didn’t automatically start even though it was set to

I clicked START and all was well.  I will do more troubleshooting on this the next time I need to reboot the Hosts.  I’d rather not evacuate all my Guest VMs for this reason alone. For now I’m happy being up to date.

I noticed when i did some maintenance and rebooted the ESX hosts and i could not get High Availability (HA) back up and running.  I kept receiving the error:

An error occurred during configuration of the HA agent on the Host
Enable firewall failed: vim.fault.HostConfigFault

Strange.  Looks like neither ESX hosts are talking to each other.  VMWare communities suggest i run:

[root@syr-nav-esx-03 root]# esxcfg-firewall -q aamclient

And it gave me:

Can’t locate XML/DOM.pm in @INC (@INC contains: /usr/lib/vmware/esx-perl/perl5/site_perl/5.8.0 /opt/ActivePerl-5.8/site/lib /opt/ActivePerl-5.8/lib .) at /usr/sbin/esxcfg-firewall line 21.
BEGIN failed–compilation aborted at /usr/sbin/esxcfg-firewall line 21.


Turns out the problem is from the changes Asigra Televaulting DS-Client wanted me to make to each ESX host in order to get host based backups working.  The DS-Client requires Perl to be upgraded to at least 5.8.1. And in order to make sure ESX continues to function after the Perl upgrade they states on page 419 of dsuser.pdf:

After the Perl upgrade, the vmware-cmd command may no longer work since it is not compatible
with the new Perl. You will, however, need this command in order to register restored virtual
machines. To fix this problem, go to the /usr/bin/ directory and edit the vmware-cmd file as well as
the vmware*.pl files in this directory; change the first line in each file to refer to the old Perl
installation (replace #!/usr/bin/perl with #!/usr/bin/perl.old). Run vmware-cmd without
parameters to verify that it works.

So i guess they forget to mention /usr/bin/esxcfg-firewall also needs it’s perl path changed to /usr/sbin/perl.old.  This makes me wonder what else migth brake from the Perl upgrade.

My approach to backing up guest VM’s from ESX3.5 is two-headed:

• Backup the data like you would a normal machine by attaching via SMB, NFS, SQL, etc.  Just get the data
• Backup the Guest VM via the ESX host (pull a bare metal snapshot of the entire system)

Why backup everything twice you ask?

Well fortunately for you and unfortunately for me I have a real world experience that will help answer this question. I had a vm guest failure and had to rebuilt a vm then restore data and settings.  Suffice to say this took forever. So here’s why you can do both…

• If data is deleted you have backup sets that have just pure data – files that are used by applications and users. If a single file gets deleted, corrupted, or anything else _bad_ the restore for this file(s) is quick and easy.  Restoration of this file(s) doesn’t need to affect every other user connected to the system in question.
• If the OS becomes unusable, system files fail, vmdk (virtual disk) files on the host get corrupted then you now have the full machine backup to simply turn back on from date X/Y/Z. Asigra actually takes a native ESX snapshot at the time of backup.

Using Asigra Televaulting (or a small handful of other backup systems) allows us to perform backups that are compressed and bit level.  So, if I have a VM that has a 20GB vmdk disk, but really only has 8GB of data in the guest system, then the backups will actually be less than 8 because all that free space will be compressed down to nothing and then the 8GB of real data will further be compressed down.

My schedules are now setup to backup real data each night, but backup the OS (again using the native ESX snapshot) once a week.  For the most part, system settings and configurations aren’t happening each day. So, if the guest VM dies, I can simply restore the VM to the last weekly backup (as of at the most 6 days) then restore the real data to to that machine (as of at the most the night before).

• A note for Asigra users – I’ve had a lot of problems backing up via VI.  So far attaching directly to the ESX hosts seems to be working great.  The one fallback for this is you need to setup rules to ensure that VM’s stay on the same ESX host.  Asigra only knows that a guest VM is on the host you originally configured it to be backed up from.

So I was having some pretty significant performance problems with a vm running Windows Server 2003.  I thought it might be due to the fact that I had so many snapshots.  Fellow ESX admins over at the VMWare communities confirmed this to be the case.

So my next step would be to combine all the snapshots and get rid of my delta’s by committing them all. I was going to run some guest updates first, so I again as always, made a snapshot. Something either guest or host related happened and the guest powered off.  When I went to turn it back on VI complained:

too many levels of redo logs

uh oh! the guest wouldn’t turn on!  it turns out that this snapshot was in fact the 35th snapshot taken for this machine. This 35th snapshot didn’t complete correctly and was corrupted. it also turns out there is a 32 snapshot limit for VM guests.  Good to know VI tells you about this! <sarcasm included>

After frantically searching the web and forums for a solutions someone pointed me in the direction of a post here: http://zealkabi.blogspot.com/2008/10/virtualcenter-shows-no-snapshot-but-it.html which clearly shows the process i need to use to commit my snapshots, specifically Solution B:

If solution A did not work then next step to follow is: use vmkfstools -i to consolidate snapshots.
1. You can export the disk with vmkfstools to recreate the virtual machine:
2. Execute the following command to create a directory for the new disk:\
# mkdir /vmfs/volumes/UUID/new_RHEL5
3. Execute the following command to point vmkfstools at the last snapshot file:
# vmkfstools -i RHEL5-000001.vmdk /vmfs/volumes/UUID/new_RHEL5/new_RHEL5.vmdk

Three hours later, snapshots 32 through 1 committed and a single vmdk.  While this commit process was running I realized i could have simply told VI to run vmdisk00032.vmdk instead of the final (and corrupt) vmdisk00035.vmdk.  this would have been the quick resolution to get me back up and running, and I could have don the snapshot committal at a better time.

Huge thanks to patrickds from the VMWare communities and SANJAT KABI (http://zealkabi.blogspot.com) for their knowledge!

Just noticed this syslog appliance was recently released. its free and runs rsyslog and phplogcon for the web interface. its pretty powerful, and it comes working out of the box…oh ya and it’s free.

One note – GD seems to be broken in the current release. its as simple as “sudo apt-get install php5-gd” then “sudo /etc/init.d/apache2 restart”

Syslog Appliance