Taking the plunge...

I've lost count of how many time I've heard people say "I installed Vista, but I'm back to XP now". People who know enough to be comfortable with rebuilding their own systems of course, one wonders how many of the general population would do the same if they knew how.

I'm going one step further now... today I received a free VMware Workstation 6 key (being a VCP and all), and I redeemed it for the Linux version of Workstation. I'll keep a disk around with XP on it for when Team Fortress 2 comes out, but otherwise I'm sure any other app I'd care to use (that doesn't have a Linux equivalent) will run fine within a VM... Visual Studio 2005 is about the only one I can think of in that category to be honest!

VMware Workstation 6.0 finally released....

Yes sir, the latest major revision of VMware Workstation has finally gone gold. The revision number is not far above the RC, so I'm guessing there haven't been many changes between the RC and the final product. I found the RC to be rock solid, and was running the various betas for quite some time (and submitted my share of bugs, which were fixed over the course of the beta). Go grab it now!

Epitome of Wasting Time...

Say goodbye to your precious time...

Desktop Tower Defense

Microsoft Virtualisation MBSM getting into gear...

That's right, the Microsoft MBSM (Marketing Bull Shit Machine) is really cranking into gear, this time via Mike Neil posting on the otherwise respectable Windows Server Division blog (Ward, none of this is directed at you, you do a fine job. It's when those marketing types somehow take over your blog that things go down the toilet).

Seriously Mike, do you think the ppl reading Ward's blog are managerial types who don't see straight through marketing BS like "We’re designing Windows Server virtualization to scale up to 64 processors, which I’m proud to say is something no other vendor’s product supports."

You're proud to say you're DESIGNING something that no other vendor's product (currently) supports!?!?! Come back when you have a shipping product that comes somewhere near what your competitors are (currently) doing in terms of performance, scalability, management and stability, then you can boast about what features you have that they don't.

I'd also be interested to know where the requirements came from to develop such a feature... applications requiring 64 processors are generally not the kind of thing you would virtualise... how would you go about allowing for hardware failure with such a VM? Have another 64way piece of hardware sitting there doing nothing, that you can fire up in the event of a hardware failure on the first one? Gee, we're back with the physical world problem again!

Seriously, stuff like this just makes me think Microsoft aren't on the right wavelength. Another blog somewhere once stated that while everyone currently in the virtualisation market has solved the physical problem, only VMware has actually solved the VIRTUAL problem, by way of their management software and Vmotion. It's pointless to tout things like hot add NICs and CPU's, how many physical servers will let you hot swap a failed CPU? When was the last time you were in a datacenter and tried to do that? How about trying to do that on a blade? What would happen to the instructions the VM was executing at the time the CPU failed - are they gona seamlessly failover to a different CPU? How many applications that require going from 2 to 4 CPUs and then back again are really suitable for virtualisation? That kind of issue is generally solved by scaling out the number of machines running the application rather than increasing the CPU available to a single machine. What happens to the HAL and the kernel when you try to go from 1 CPU to 2 CPU's and then back again? I can only assume the enlightenments in Longhorn will solve that issue...

I hope when the product launches it has something like Vmotion, otherwise there's no way it can hope to compete. Microsoft is already on the back foot in terms of the platforms that will be supported on top of Windows Virtualization, and one of Microsoft's most successful arguments against Linux is "cost of acquisition is a negligible part of TOC". Do they hope to fly in the face of that by expecting if they give Windows Virtualization away, it will be a convincing enough argument to switch?

Time to buy some shares in VMware. I can see the virtualisation market playing out much the same way as search currently is.

VMware Workstation 6.0 hits RC...

Yes that's right, build 42757 was released on the 23rd and is the first public RC of Workstation 6.0. Aside from the usual slew of bugfixes, the biggest deal for anyone using it is that debugging is finally configurable! In the betas, debugging was set to Full and there was nothing you could do about it, which is reasonable enough for a beta I guess. Now it's set to 'none' by default :-)

It's also a whopping 280MB download... the current version of Workstation 5.x is only ~90MB!

Windows Server 2003 SP2 released!

Pick your flavour, and check out the deployment guide, a useful read if you're gonna slipstream R2 with it's 2 CD's. Here's hoping Microsoft never do that again...

The Virtual garbage just keeps coming...

First of all was VMware's dummy spit regarding Microsoft's so called 'anti-competitive' behaviour (in a product that doesn't even exist yet), now Microsoft are starting to crank up the marketing machine for their upcoming virtualisation product. Seriously, when are these companies ever going to learn that you do not win the hearts and minds of the tech community with this kind of tripe?

VMware should be thanking Microsoft for creating the market that allows VMware to exist. Microsoft created / stole the software and had the marketing nouse and questionable business practices that allowed the inexpensive x86 architecture to proliferate (inexpensive compared to the other proprietary Unix options available at the time - Linux was little more than a hobbyist OS in 95). Microsoft created the platform that was so easy to program for, allowing many companies to develop applications. That platform was so easy to code for and at the same time so forgiving, many, many, MANY crappy applications were created and installed on servers, leading to the side by side incompatibility problems that virtualisation solves. The same problem that creates the massive under-utilisation problem that virtualisation perfectly solves. So please, VMware, a little credit where credit is due.

Additionally, VMware should realise by now where their bread is buttered. Windows is the mostly widely virtualised OS, and who is driving that virtulisation? Shock horror - it's the Windows engineers and admins! So just who do you think you are winning over by whining about Microsoft? Not the majority of your customer base. The Linux guys have already forgotten about you with XEN and KVM on the move, and the Unix and big iron guys have been doing this stuff for ages anyway.

But this latest creation from the marketing people at Microsoft just blows my mind. Do they think we are stupid? There is just something about making claims about a product that is 12 months away from RTM that really grates me. Like the line about having 8-way virtual machines, "a feature our competitor doesn't have yet". Oh, really? Well how about a *little* feature your competitor does have that you do not... A ROCK SOLID, WIDELY DEPLOYED ENTERPRISE VIRTUALISATION PLATFORM THAT YOU CAN ACTUALLY BUY RIGHT NOW!!! FFS Microsoft you idiots, what is the point in claiming a '1 up' over VMware with (1) a feature in a product that isn't even in public beta and (2) how many people do you think are going to virtualise an application that requires 8 CPU's!!! There are numerous other one liners throughout this video, like the one about interoperability. Wow, you can certifiably run SUSE on the Microsoft product! In fact you can prety much run anything... oh wait, no sorry, you can't run NT 4. Or anything 16 bit. Legacy Novell, Solaris, Red Hat, BSD... ummm, no, no, nope and no again. This video says one thing that is correct - customers want a virtualisation platform that they can standardise on. And it won't be Microsoft's, because you won't have a virtualisation platform that supports the widest range of guests. I also love the way it also glosses over the VMotion feature... "oh yeah we'll have the by the time the product ships." No demo though.

Imagine what the Microsoft product will be capable of in 3 years time. Now imagine that they shipped the Longhorn generation of virtualisation 3 years ago. That's where VMware are at this very point in time. I can't remember who said it where, but right now Microsoft have a solution for the physical server problem. But currently VMware have a solution for the virtual server problem, by way of VMotion and VirtualCenter. And who knows what they have in the pipeline. I'm sure that before Longhorn ships, VMware will have 8 way capable VM's, hot add everything, true high availability (ie underlying host goes down and guests pause for a few seconds then just keep running), and even better management options. They're already way ahead with the VirtualCenter web service - I somehow doubt Microsoft will be embracing the openness and flexibility that web services allows... someone might write a better management interface than Microsoft (and if you've seen the beta for System Center Virtual Machine Manager, you know that wouldn't be hard at this point in time).

I can't imagine what other propaganda will be crammed down our throats from both sides in the year to come. One can only hope that VMware takes the Microsoft FUD in their stride rather than trying to spread their own, and act with the maturity that their product has.

About bloody time!

With the aging of the population, and more people who were brought up with consoles and PC's in the household moving into positions of influence and amassing majority in the academic and media circles, I am sure studies like this will start showing up more and more.

Let's just hope each successive generation doesn't fall into the same traps as the last. You'd think the baby boomers who were raised in an era of moral outrage against rock music wouldn't have such short memories.

Attack of the Beta's! Longhorn February CTP released...

Build 6001-16461-070209-1510 (yowza!) is out now... I'm guessing this is not a milestone release, but will probably grab the bits anyway and waste some time seeing if I can install IIS 7 on 2003 and trying to find a mention of the final product name!

New Workstation 6.0 BETA build out...

Build 39849 of the Workstation 6.0 beta is out... and boy am I loving this product. It's so good, even the free price tag of Virtual PC 2007 is not worth it in comparison.

Go get it now!

Bullshit Vendor Whitepapers...

Mike Laverick over at RTFM (link in my linkage section) has made a post regarding a Dell whitepaper that has just been released, that purports to compare "2 socket vs 4 socket hardware", specifically the Dell 2950 vs the HP DL585 G2. As a longtime reader of Mike's blog and someone with enormous respect for his opinions and contribution to the VMware community, Mike - I'm severely disappointed in you for simply reporting this as if you were on Dell's payroll. These kind of whitepapers prove absolutely fuck all, and are nothing short of pure marketing bullshit.

I guess anyone reading the title of the paper, the 'Executive Summary', the 'Introduction' and then glossing over the actual hardware specs used in the testing and going straight to the results numbers would reach the same conclusion... 3 x 2 socket 2950's show more grunt for less cash than 2 x 4 socket DL585 G2's.

First, comparing retail price is useless - anyone who has worked in a large enterprise knows how close Dell, HP and IBM really are in terms of price. Second, they are comparing Intel in the 2950 vs AMD in the 585 G2. But third and worst of all, the 2 socket Dell boxes are loaded up with quad core CPU's, where the 4 socket HP's only have dual core CPU's!!! Gee, what an 'Apples vs Apples' comparison!

I'd probably be less annoyed if the authors didn't go at such lengths to hide what is actually being compared. Count how many times they refer to sockets, and how many times they refer to cores. Look at where and how they actually say "we had 4 cores per socket in our stuff and only 2 cores per socket in their stuff". How about this direct quote

... this paper compares three two-socket Dell™ PowerEdge™ 2950s with two four-socket Hewlett-Packard™ ProLiant DL585 G2s. Or more simply – 6 sockets vs 8 sockets.

What, saying 24 cores vs 16 cores instead of 6 sockets vs 8 sockets is more complex? Oh, wait... no, that would screw their results.

And that's the funny part - going on cores, the performance numbers actually work against Dell. I quote:

...these tests show that three PowerEdge 2950 two-socket servers can provide up to 44% more performance...than two HP ProLiant DL585
G2 four-socket servers.

So with 50% more cores than the HP hardware, Dell gets 44% more perfomance... ummm, either my math sucks, or your boxes suck Dell.

*UPDATE* Good pickup by Kharbin over on the VMTN forums... the Dell boxes also have half the RAM of the HP boxes, if you bring them up to the same level of RAM, the prices are pretty much the same too!

OK, maybe that's being a little harsh, and I know that Dell are not the only offenders when it comes to psuedo-research papers, but this kind of thing really pisses me off. There are so many permutations that would have made this whitepaper meaningful. Why didn't they use Dell 6950's instead of HP 585's? Why didn't they at least use a HP DL580 G4 (or even better a Dell 6850) thus making it an Intel vs Intel comparison? Why didn't they use an equal amount of cores for comaprison (ie 2 dual socket quad core Dell's instead of 3)? How would 4 single socket quad cores stack up against 2 dual socket quad cores? You get the idea.

The whole scale up vs scale out question with regards to VMware is something we'd all really like to see some solid numbers on, but to date I don't think any hardware vendor has done the work. Maybe the fact that a 4 socket box is often more than twice as expensive as 2 x 2 socket boxes has something to do with that. The hardware manufacturers are already staring down the barrel of severely diminishing sales thanks to virtualisation, better for them to get the higher margins on 4 socket boxes I imagine. So rather than Dell shoot themselves in the foot by saying 3 x 2 socket Dells are better than 2 x 4 socket Dells (or god forbid, 2 x 2 socket quad core Dells are roughly equivalent in performance to 2 x 4 socket dual core Dells, but the 2 x 2 socket boxes are 200% cheaper), I guess the only choice these pseudo-researchers had was to compare them to another company's 4 socket product. And I guess Dell is really hurting since HP knocked them off the #1 PC seller spot.

I can only hope that HP retort when the quad core Opteron's are released. No doubt Dell would cry foul if that happened though, so I'm hoping for at least a Dell 2950 with Intel quad cores vs a HP DL385 with AMD quad cores (the 385 is HP's 2 socket AMD box for those who aren't familiar with HP). I'm sure you can guess what my money will be on.

And just to be clear, I'm not arguing that 'few big boxes' is better than 'many small boxes' (what design you use depends on the risk aversion of the stakeholders / customer more than anything) - all I'm saying is that this paper needs to called out for what it is... another BVW &trade

VMware Workstation 6.0, totally awesome!

I've made the switch to the Workstation 6.0 beta on my main desktop now... it's totally awesome. Although I did find a few bugs, they are mostly minor or cosmetic (except for one really huge, massive one which I'll describe shortly). And sure the full debugging mode might have a little impact on performance, but not enough to make me care. The fact that I can use VNC to connect to VM's now (just by ticking a box in the VM's options - no software is required in the VM, not even VMware tools) saves some memory to compensate for that anyway - running a VM with the Workstation GUI uses around 23MB RAM and around 1% CPU per VM, whereas connecting to a VM via a VNC Viewer uses around 4MB RAM and pretty much 0% CPU. Since I usually run 3-4 VM's a time, that's a big benefit... even though I may need 4 VM's running, I rarely need more than 1 console displayed. Plus there's just something cool about watching a machine boot via a VNC window... that's right, since the VNC viewer is connecting to vmware-vmx.exe on the host rather than a VNC server in the guest, you get a proper console connection like an iLO on a HP server.

Then there's all the other numerous tweaks... mounting vmdk's from the GUI, the dual monitor support (bloody awesome)... it's going to be a truly great product. I don't care if Virtual PC 2007 is going to be free - like so many things in life, you get what you pay for! Although being a VCP, I do get Workstation for free anyway, but you know what I mean :-D

Now about that big nasty bug. If you take a snapshot of a VM, then add a pre-created virtual disk, then later revert to the snapshot... the vmdk file is deleted from disk!!!. I was expecting that the file would just be removed from the VM itself (ie removed from the .vmx file), so was quite surprised (and alarmed) to see the vmdk disappear from the filesystem. I have filed an SR with VMware (they use their support request system for bug tracking, oddly enough), and anyone reading this should do the same!

WSS 3.0 unattended setup...

WSS 3.0 totally rocks, and to facilitate faster dev environment provisioning, I whipped up a quick unattended install that y'all might be interested in.

It assumes you have already install IIS and .NET Framework 2.0, and enabled the ASP.NET 2.0 web extension (aspnet_regiis -i -enable). So for the rest, I have an ISO with 2 directories, NETFX30 and WSS30. Then I have the following in a batchfile in the root of the ISO:

@echo off
::
:: Install .NET Framework 3.0
echo Installing .NET Framework 3.0...
echo.
cd NETFX30
start /wait dotnetfx3.exe /q /norestart
::
:: Install WSS 3.0
echo Installing WSS 3.0...
echo.
cd ..\WSS30
xcopy config.xml %WINDIR%
start /wait setup.exe /config %WINDIR%\config.xml
::
:: Sleep 10 mins
echo Sleeping for 10 mins for WSS config to complete...
echo.
sleep 600
::
:: Reboot
echo Rebooting...
echo.
shutdown -r -t 10

And the the config.xml file has the following:

[Configuration]
[Package Id="sts"]
[Setting Id="REBOOT" Value="ReallySuppress" /]
[Setting Id="SETUPTYPE" Value="CLEAN_INSTALL" /]
[/Package]
[DATADIR Value="D:\WSSv3\Data" /]
[Logging Type="verbose" Path="%temp%" Template="Microsoft Windows SharePoint Services 3.0 Setup(*).log" /]
[Setting Id="SERVERROLE" Value="SINGLESERVER" /]
[Setting Id="UsingUIInstallMode" Value="0" /]
[Display Level="basic" AcceptEula="yes" CompletionNotice="no" /]
[/Configuration]

But of course all those square brackets should be angle brackets. And obviously all my boxes have a D: drive. If you already had a farm, you would change the SERVERROLE value to "WFE" (Web Front End).

The unattended setup of WSS 3.0 is pretty horrible... the SQL Embedded instance does a crapload of configuration long after the actual setup completes, hence the 10 minute sleep I have in there. If you reboot before that configuration finishes, the WSS 3 config wizard will give you a world of hurt. If you want to modify it to reboot exactly after the configuration is successful, look for Event ID 101 from source "SharePoint Products and Technologies Configuration Wizard" in the app event log.

But besides that, the product is awesome!

Just show me the damn clock!

One of the things that has always really annoyed me is the systray clock... so easy to turn on in the UI, oh so hard to do in an SOE! Well not anymore... this command will do it:

REG ADD "HKCU\Software\Microsoft\Windows\CurrentVersion\Explorer\
StuckRects2" /v Settings /t REG_BINARY /d
"28000000ffffffff02000000030000006b00000020000000feffffffe
20300000205000002040000" /F

*NOTE* All that should obviously be one line... just found a bug in the .css of this site that I'll need to take care of later!

Of course, if you want to make it stick for all users, mount the Default User's NTUSER.DAT and make the change there too. No, changing the .DEFAULT hive won't work - that is the "All Users" ntuser.dat... stupid naming I know, but the actual "Default User" registry hive isn't exposed by default. And since a new user profile gets whatever is in the "Default User" hive, you need to modify it (in any SOE I have had anything to do with, I always make user related reg entries in both All User and Default User, just to be sure)

Happy New Year!

That last year seemed to go fast when I look back at the blog entries, but when I think about the year I feel a mild sense of accomplishment which is cool. Hopefully this time next year I'll have a great sense of accomplishment... got some stuff planned for this year, here's hoping it all works out for me, and for everyone. Happy New Year!