I’m always in search for better ways to use Outlook more efficiently. My coworker recently brought to my attention a neat little tool designed to work within Outlook called Xobni. It’s main purpose is to network your contacts and messages with outlook into one consolidated and easy to use sidebar within Outlook. I downloaded this tool for free (the product is currently in public beta) and have been pleasantly surprised with it. For being beta, installation went extremely smooth and it immediately indexes all messages in my Outlook and all my attached PSTs. The built in sidebar provides you with a quick snapshot of messages associated with the contact that you’ve currently highlighted. The ability to drill down into the network has already paid off for me in the past 2 days. My only feature request have so far is to integrate saved IM conversations into the network that is created within Outlook. Give it look, I highly recommend this tool.
Here’s a basic script for applying patches on ESX 3.0.X systems. This has not been tested on ESX 3.5 because Patch Management is supposed to be built into VC. I will update once I am able to test the script on 3.5. I recommend that no VMs are running on the ESX while you run this script.
Usage
- Install all patch packages into /var/updates directory
- Create script file in /var/updates directory
- Execute script
Code
!/bin/bash
echo Installing ESX patches…
for i in ESX-*.tgz
do
tar -xvzf $i
done
rm -f ESX-*.tgz
for j in ESX-*
do
cd $j
esxupdate –noreboot update
cd ..
done
rm -r -f ESX-*
grep -i TOTALS: /var/log/vmware/esxupdate.log
echo Installs are complete. Please Reboot
If you have not heard by now, Novell has announced that they will acquire a company called PlateSpin. For those of you not familiar with PlateSpin, they are a software company that specializes in virtualization monitoring and virtualization portability through their main products, PowerRecon and Power Convert. Personally I have worked very closely with PlateSpin, integrating their PowerConvert software into about 90% of my clients’ “lights out” disaster recovery strategies . Whether it be a virtual to virtual or physical to virtual migration plans, PlateSpin’s PowerConvert fits nicely into most companies’ business requirements for disaster recovery. PowerConvert is not perfect and might not be a true fit for all clients, but that’s for another entry.
Whenever one of the big guys gobbles up a small software company, the first question anyone asks is “who benefits from this”? First and foremost, this benefits Novell by giving them a proven multidimensional product suite to build their SUSE Xen product line around. Secondly, this acquisition gives Novell access to PlateSpin’s client base which is comprised primarily of VMWare users. PlateSpin does support a host of other virtualization platforms including VirtualIron and VirtualPC, but the reality is most PlateSpin clients run VMWare. This will be huge in the coming the years as virtualization technologies other then VMWare begin to close the gap in terms of functionality and reliability. The PowerConvert specializes in converting virtual machines from one platform to another. Novell will have a proven and mature tool that will easily move customers away from VMWare. And finally, this acquisition is another shot at VMWare’s dominance in the marketplace by going after the functionality that VMWare is tyring to bring to market, namely their Site Recovery functionality in the next release of the ESX 3.5 platform.
It will be interesting to see how Novell handles this acquisition and only time will tell if this is a right move for them.
When installing VMWare Lab Manager 2.5, the first thing that runs is a validation check of your system to ensure that the installation will go smoothly. If any errors or warnings are present, the installation will not continue. In certain testing scenarios, you may want to bypass these checks. Here is the command to do that:
[setup filename].exe /v”SUPRESSPREREQ=TRUE”
Screenshot of warnings prior to installation of VMWare Lab Manager
Screenshot of command line to suppress the validation checks
I have not given up on my quest to simulate latency using the the built in LAN configations of VMWare’s workstation. In researching the problem, I noticed that VMWare had released a new build of the workstation, 6.0.1 -build-55017.
Here is the complete list of the updates for 6.0.1:
http://www.vmware.com/support/ws6/doc/releasenotes_ws6.html#601
Other then the fact that no latency issue is mentioned, it’s interesting they now support the upcoming Longhorn OS. They’ve also added the same support tool that they have in the VI Client to dump troubleshooting data to be sent off to support. Please keep in mind this tool collects some basic information about the system the workstation is running, including services and processing running, all the event logs, and the VMWare Workstation licensing information.
Back to the troubleshooting….
I also tried a few other steps to help troubleshoot this scenerio.
- Upgraded to 6.01
- Upgraded Windows VM’s vmtools to latest version
- Installed a second windows server 2003 VM to remove anything possible network issues with the linux VM and windows VM.
There have been no changes with the 6.0.1 build. I am still seeing the same results as my previous post. My next step is the run this on brand new install as my current system image is suspect.
While working on another blog entry involving latency thresholds with ESX and VirtualCenter, I came across a disturbing issue with VMWare workstation and their built in packet loss emulator. I came to the conclusion that it does not simulate any packet loss no matter what % you set it at.
Environment
- VMWare Workstation 6.0.0 build-45731
- VM Team with LAN of 1Kbps and 99% packet loss
- 2 Virtual Machines attached to the internal team LAN
- VM 1: Windows Server 2003
- VM 2: ESX 3.0.2
Results
Screenshot of the LAN settings with VMWare Workstation
Screenshot of the ping test from VM 1 to VM 2
As you can see, there really isn’t much packet loss or latency going on here. I still need to confirm this issue on another system before I completely thrown in the towel. Sadly, this is a feature I was hoping to leverage in order to setup my test environment for next blog. Now I will have to build my laptop on my Network Nightmare device. Hopefully, the next build will correct this problem.
Virtualizing Exchange is a discussion I have had with almost every customer I’m in front of. Some of the customers are open to virtualization of Exchange while others are completely against it. For the most part, the debate about using VMWare in production has pretty much subsided in the favor of VMWare. Exchange is the last great debate.
In my personal opinion, I think most SMB customers could benefit from the virtualizing Exchange without impacting performance. In most instances, a lot of customers have only 1 exchange node running with very little in the way of DR. VMWare helps them build some HA functionality where it didn’t exist before. That’s my quick 2 cents.
The following is great document from VMWare and Dell.
http://www.vmware.com/pdf/Virtualizing_Exchange2003.pdf
As I’m working on my first recap post of EMC World, my buddy has posted an excellent ( and at times humorous) post regarding NBU best practices. Please have a look.