Bringing it back, baby!

But seriously, if you have not heard about the Tech Field Days run by Stephen Foskett, you’re in for a treat. This will be my second time attending a Tech Field Day, having attended Tech Field Day #8 as a delegate last year. Now that I have a much better idea what it’s all about, I wanted to shine some attention on the event in hopes of increasing awareness and, well, evangelizing it a bit.

Full Disclosure

The entire trip is paid for by the sponsors indirectly (Tech Field Day handles the payments on my behalf). Sometimes there are branded swag gifts or product demos given out. Nothing is expected or required. I wrote about some of the sponsors, and not others, during my last visit. I was invited back – I guess that proves that you don’t have to go bonkers with posts or cheerleading for any of the sponsors.

What Sets Tech Field Day Apart

First off, it’s an incredibly fun event. Don’t get me wrong – the days are grueling and I typically got less than 4 hours of sleep a night, but that’s because I spent so much time learning, writing, and conversing with the presenters (such as The Most Interesting Man in the World - stay Cloudy my friends!) and fellow delegates until the wee hours of the morning. There are no censors, NDAs, or filters at Tech Field Day – Stephen is adamant on making sure your voice is heard, even if you hated the presentation.

The Open Flow of Communication

Sponsors of Tech Field Day are prepped before hand to understand that they are delivering content and opening themselves up to exploration from technical engineers. Delegates aren’t writing press releases, they are analyzing the product or service and picking it a part. That doesn’t mean dropping F-bombs and setting fire to the carpet, but if they want to express the negatives of a product, delegates can and are encouraged to!

Generally frowned upon

Additionally, the presentations are typically not death by PowerPoint (or at least, the GOOD ones are not). Technical demos with live products, whiteboards, and tours are common place for a Tech Field Day. The idea is to open a dialogue with you (the delegates) and them (the sponsors) to get unfiltered feedback. Thus, opinions matter!

This is roughly what it looks like at a Tech Field Day presentation. This is from a session at TFD8 from Pure Storage (the SSD whisperers).

With all that said, the sponsors and content are typically top notch and bring their A-game to the presentation. The discussion is focused on something that’s interesting, sometimes in stealth mode and coming into the market as a new idea, and therefore fun to learn about and engage with the designers.

Social Media Fiesta

During a presentation, social media is going nuts. Hashtags in Twitter are flying all over the place, blogs are being updated, and a video camera is recording (and live streaming) everything that’s going on. It’s a veritable waterfall of information pouring out of every session. The Tech Field Day staff do a great job at making sure all of the delegate content is quickly spread to the world, and keep the participants updated on the various streams and events taking place.

The fail whale is often caused by all of the delegate commentary

Becoming a Part of Tech Field Day

I encouraging others out there who are of an independent mind to throw your hat in the ring. They are always looking for new blood. Or, if you know someone who would benefit from a Tech Field Day, nominate them here.

You can also nominate sponsors that you’d like to see present at Tech Field Day.

This is kinda what it is like to deal with Microsoft licensing

Who is Affected?

This issue doesn’t specifically stop at VMware’s vMotion. Seemingly any Windows server license falls under the agreement umbrella. It’s rather sad that in this day and age we’re still having to jump through hoops around licensing, although I would imagine profit margins come in to play here somewhere. I find it especially confusing since Microsoft has a product that is also hurt by this policy, and an ecosystem of overt complexity only breeds FUD (Fear, Uncertainty, Doubt).

Tasty and delicious

Working Through the Pain

I’ve always been a strong advocate of the Windows Data Center level licensing, as it seems to be extremely friendly to virtual environments. In a nutshell, Data Center is sold on a per socket quantity. Any Windows server using that socket is licensed and can use whatever edition needed. The price is typically competitive compared to the Enterprise license, which only allows for 4 guest VMs to be created per license. This doesn’t mean that you have to install Windows server Data Center edition (typically overkill for most people) in the virtual machines.

As an example, a two socket VMware host can run multiple copies of Windows server in all flavors (Standard, Enterprise, or Data Center) as long as two licenses of Windows Data Center are purchased.

Unless you already have a significant investment in lower level licensing, such as a server farm running non-OEM’ed Standard licenses, go with Data Center.

Note: I specifically leave out anyone with an enterprise agreement with software assurance. I imagine that at that level, you aren’t worrying too much about licensing.

Big Brother Isn’t Watching You

Not that I’m advocating violating your license agreement, but to be clear there is no call home feature that I’m aware of that sends the five-O to your business. Definitely know what you’re getting in to before you crank up DRS to fully automated and let it play shuffleboard with your VMs, but don’t worry about events like HA restarting your VMs due to hardware failure (perfectly within your rights of the licensing).

If you are audited, though, you may be in for some fun. And by fun, I mean not fun.

However, that does you no good if you’re on the other side of the glass of PowerCLI utopia. Sometimes the hardest part to getting started is figuring out the first step. So, this post will go over some suggestions on how to set up your workstation to form a PowerCLI zen garden and some great resources (both online and in dead tree form) to help walk the path to enlightenment. Once you have a basic level understanding of how this scripting language works, mental sanity will be restored and a guy from Western Union will show up with a seventy year old letter stating that Doc is still alive and living in the wild west of 1885.

You don’t have to wear this to learn PowerCLI, but it helps

Go Grab PowerCLI!

Make sure you have PowerCLI installed before we go any further. So, go download and then install it. Preferably in that order.

We’re not actually going to use the “stock” version of PowerCLI in this tutorial so don’t worry about launching any applications after you’ve finished installing.

There’s just no sizzle with the stock editor.

PowerGUI – 1.21 Giga Watts of Power

This section could reasonably be entitled “use a scripting editor” but, quite frankly, I think PowerGUI is a great tool and highly advocate its use. The free version is amazingly well written with lots of community support and rich with features.

The main thing I like about PowerGUI is that it’s extremely stable. I’ve used a number of different GUI editors and have been disappointed with bugs or “flashy graphics” that offer no value. For the money (free) this is just one of the best. Even with a 60 second auto save built in and my habit of saving constantly, there’s nothing worse than having your editor puke in the middle of a flash of code to really put a rain cloud over your head.

Also, like any good editor, this one offers a nice set of window panes for variables / console / editing, auto-complete for cmdlets, and coloration of the code. There’s a whole slew of other features, such as PowerPacks and tabs, but these are the meat and potatoes.

Meat, Potatoes … and scripting.

Research Material and Rockstars

There are a number of great resources available online to help you learn PowerCLI. My main source when I was learning consisted of the three rockstars of scripting: William Lam (Virtually Ghetto), Luc Dekens (LucD Notes), and Alan Renouf (Virtu-Al.net). Very humble guys who donate a lot of their time to evangelize for the community and help people solve scripting road blocks on their websites, VMTN forums, via twitter, and who knows else where. I highly recommend following them on twitter and reading their blogs as a great way to start.

Additionally, Alan and Luc wrote a great book on the subject called “VMware vSphere PowerCLI Reference: Automating vSphere Administration“ (Amazon) that I own and use frequently. I call it the PowerCLI bible. Go get it, it’s like 20-something bucks. Easily worth it.

A treasure of scripting knowledge is behind this cover

Learning to Script

Aside from formal course work, the best advice I can give is to snag other scripts and reverse engineer them. The book shown above has a lot of example scripts and explains how they work. Use this is a blueprint to begin making your own scripts. Another tip is to start simple and try to emulate many of the features that the vSphere client performs for you – you should already be familiar with how these work, so it’s just a matter of learning the language of PowerCLI. Of course, it helps when you’re not doing this in production, so make sure to pick a lab environment if you’re not 100% confident in your code (which should be the case if you’re just starting off!). The best way to firewall yourself from doing any harm is to script directly against a host that has only lab VMs on it, or a lab vCenter (if you have that luxury).

Additionally, check the PowerCLI VMTN forums for clues. Many of the scripts that you want have already been requested and fullfilled by a number of folks that want to help. I don’t advise simply plopping down a post that says “do it for me” – I typically ignore these. However, if you post your script and ask specific questions, such as “why am I getting this error” or “what am I missing” you should get plenty of help.

If you can find a mentor in your office or on twitter, that’s a great start as well.

This is my scripting mentor, Battle Owl.

Scripting Exercises

Try to figure out the code for the following activities. The answers are below in the next section.

1. Power on a VM
2. Gracefully shutdown a VM
3. Take a snapshot of a VM
4. Figure out how many VMs are in a Cluster
5. Rename a VM

Scripting Exercise Answers

Power on a VM

Start-VM -VM "name"

Gracefully shutdown a VM

Shutdown-VMGuest -VM "name"

Take a snapshot of a VM

New-Snapshot -Name "WahlNetwork Snapshot" -VM "name"

Figure out how many VMs are in a Cluster

(Get-VM -Location (Get-Cluster "cluster name")).count

Rename a VM

Set-VM -VM "name" -Name "Wahl Network VM"

Thoughts

As you can see from the answers, much of the code is just plain English. Nothing to be afraid of! I hope this begins your journey towards using (and loving) PowerCLI.

In this post, I’ll go into more detail on this process, and then show you a trick using ADSI edit to change the location, along with the results. Comments especially welcome; I’d love to hear any pain points you’ve had with VM Folders (if any) and how you overcame them.

Note! Use of ADSI edit may not be strictly supported by VMware. The post below was performed in a home lab environment. Use at your own risk – wear a helmet!

Examining View Pool Settings

For this scenario, I’m using an automated pool in my home lab that is creatively called Home. You can see that the virtual machine settings include a VM folder set to the path of:

/WahlNetwork (datacenter name) /vm /View (parent folder) /Home (child folder)

Here is a snapshot of the settings in the View Administrator

You can also see that the two desktops that are provisioned for this pool are sitting happily in the Home folder in vSphere.

The path to this folder is referenced as a string value to View. Meaning, you can’t change the name of any folder along the path, be that the parent(s) or the child folder. If you do, future provisioning efforts will fail citing that the folder could not be found. I don’t have enough vSphere API knowledge to specifically understand why a string path was chosen; in PowerCLI you can see that each folder has an Id value that doesn’t change based on the name.

For example, my “Home” folder has an Id value of Folder-group-v69. When I change the name to “Blah” it retains that same Id. Here is a dump of the info in PowerCLI

Name Id
---- --
Home Folder-group-v69
Blah Folder-group-v69

Drastic measures are needed. We must hack the Gibson.

This is what hacking looks like. Can you spot the garbage file?

Using ADSI Edit

Active Directory Service Interfaces Editor (ADSI Edit) is an invaluable tool that is used by administrators to do custom edits to an LDAP database. VMware View actually uses a combination of Active Directory Application Mode (ADAM) and Active Directory Lightweight Directory Services (AD LDS) to store information on the environment, meaning you can use the standard ADSI Edit tool to view and configure this information.

In this particular scenario, I have used remote desktop to connect into my View Connection Server so that I can run adsiedit.msc (the console tool). You should be able to launch it directly from the Run prompt, otherwise it should be in the Administrative Tools program folder.

Once you have ADSI Edit running, right click the ADSI Edit root object and select “Connect to…”

You’ll need to modify a few of the settings in order to connect to the View environment. These settings are:

• Name: Arbitrary, but I use “View” to remind myself what this connection is in the future.
• Connection Point: Change to “Select or type a Distinguished Name or Naming Context” and then enter dc=vdi,dc=vmware,dc=int
• Computer: Change to “Select or type a domain or server:” and put in localhost
  • Note: This assumes you are locally connected to the Connection Server

Click OK and a new entry will be added to the tree. Expand the tree until you find the “OU=Server Groups” folder. This contains the View Pools. In my scenario, there is only one pool (CN=Home). Right click on this entry and select Properties.

There are tons of properties available for the pool. The one we’re interested in is the pae-VmPath value of /WahlNetwork/vm/View/Home. This is the string of the path that View is using to identify the vSphere folder (blue folder). If, for example, I wanted to change the “Home” folder to “ZeroCool” I would make the edit to the pae-VmPath property to read:

/WahlNetwork/vm/View/ZeroCool

I’d also need to change the name of the vSphere folder to be ZeroCool as well.

Results

Now that the change has been made in both ADSI edit and in vSphere, a quick check to the View Administrator reveals that the pool is ready to go. Notice that there is no “Browse…” button next to the VM folder setting.

The ultimate sign of success is seeing a new desktop provisioned in the ZeroCool folder. Maybe I should make an AcidBurn pool as well?

Thoughts

Why is the VM Folder setting not changeable from the GUI? It seems like such a harmless thing. I’ve come across multiple environments that set up their first pool(s) and didn’t quite understand the vSphere folder piece, and then discovered it was too late to change things and had some spider web of weird folder nesting. I also like the idea of changing the folder to something that isn’t tied to the pool ID, as some of my View deployment use cases aren’t strictly for end users that don’t see vSphere (such as in pre-prod testing with application groups).

Until this is a supported, editable GUI field, I will say that the method I show above works fine and I’ve used it a few times to fix some weird vSphere folder nesting. Again, however, use at your own risk (and wear a helmet!).

Who Needs Resource Pools?

That’s a good question. You can’t even make a resource pool on a cluster unless you have DRS running. So, if your license level excludes this technology, you don’t have to worry about resource pools at all. If you are graced with the awesomeness of DRS, you may need a resource pool if you want to “weight” different types of workloads for two scenarios:

• For when memory and CPU resources become constrained on the cluster, and
• … for when a workload needs a dedicated amount of resources at all times.

Now, this isn’t to say that a resource pool is the only way to accomplish these things – you can use per VM shares and reservations. But, these values sometimes reset when a VM vMotions to another host, and frankly it’s a bit of an administrative nightmare to manage resource settings on the VMs individually. I’ll give an exception to those using some sort of script, but it would require adding some creative solutions to identifying which VMs should be set to what resource values (folders, annotations, etc.).

The only debate more epic than resource pools is the infamous “Pancake vs Waffle” debacle

I personally like resource pools and use them often in a mixed workload environment. If you don’t have the luxury of a dedicated management cluster, resource pools are an easy way to dedicate resources to your vCenter, VUM, DB, and other “virtual infrastructure management” (VIM) VMs.

Why People Fear Resource Pools

People fear resource pools because they are mysterious like a wild unicorn. Ok, maybe not that mysterious, but they are a bit wonky at first. Also, they are easy to misunderstand, and thus misuse.

Here is an infographic I’ve created that shows a typical scenario where someone has deployed a resource pool without understanding fully how they work. Look through the graphic and then we’ll discuss further.

Have you had a chance to look at the picture? Hopefully yes, so let’s cover it.

Where Did I Get Those Numbers?

Let’s start with the resource pools. You’ll notice 3 bullet points for each pool – the shares setting (high or low), the amount of shares for RAM, and the amount of shares for CPU. Here’s the math and the documentation supporting it:

• RAM is calculated like this: [Cluster RAM in MB] * [ 20 for High | 10 for Normal | 5 for Low]
• Our cluster has 100 GB of RAM (see the blue section) and thus the math is: 102,400 MB of RAM * 20 = 2,048,000 for High and 102,400 MB of RAM * 5 = 512,000 for Low

• CPU is calculated like this: [Cluster CPU Cores] * [ 2,000 for High, 1,000 for Normal, 500 for Low ]
• Our cluster has 100 CPU cores (see the blue section) and thus the math is: 100 * 2,000 = 200,000 for High and 100 * 500 = 50,000 for Low

Based on this math, the Production resource pool has roughly 80% of the shares. However, when you divide those shares for the resource pool by the number of VMs that live in the resource pool, you start to see the problem. The bottom part of the infographic shows the entitlements at a Per VM level. Dev/Test has more than twice what Production has when looking at individual VMs.

Giving Dev/Test more weight than Production? Eeeeexcelent. (Doh!)

Maintaining the Balance

The trick to keeping your resource pools balanced is to work the math backwards and never, ever use the default high, normal, and low shares values. Decide the weight of your per VM shares first. Let’s say that I want my Dev/Test VMs to receive half as much share weight as Production. Instead of worrying about VMware’s default share value calculations above, create your own. Shares are an arbitrary value that just determine weight, they aren’t a magic number.

So, let’s give Dev/Test VMs 50 shares each, and Production VMs 100 shares each. I would change the resource pools to this:

• Production would get [ 90 VMs ] * [ 100 shares ] = 9,000 shares of RAM and CPU
• Dev/Test would get [ 10 VMs ] * [ 50 shares ] = 500 shares of RAM and CPU

Much easier, right? Realistically, I could have just chosen a per VM share value of 2 for Production and 1 for Dev/Test. They would both do the same thing.

Note! If the number of VMs in the resource pool change, you’ll need to update the resource pool shares value to reflect the added VMs. Your options are to manually update the pool when the number of VMs inside change (no fun) or use … PowerCLI!

Using PowerCLI to Balance Resource Pool Shares

Now that I’ve invoked the PowerCLI SpongeBob graphic (Wooo!) let’s do some coding. This very basic script will connect to the vCenter server and cluster specified and look at the resource pools within. It then reports on the number of VMs contained within and offers to adjust the shares value based on an input you provide. It confirms before making any changes (just click or answer No if you don’t want the change).

Make sure to change line 02 and 03 to specify your vCenter and cluster name. Feel free to distribute and modify at your leisure, I take no responsibility if your environment catches fire, becomes sentient, etc.

## Variables
$vcenter = "fqdn.of.vcenter"
$cluster = "Cluster Name"

## Gather RPools
Connect-VIServer $vcenter
[array]$rpools = Get-ResourcePool -Location (Get-Cluster $cluster)
cls

## Enumerate Members of RPools
Foreach ($rpool in $rpools)
	{
	If ($rpool.name -ne "Resources")
		{
		[int]$pervmshares = Read-Host "How many shares per VM in the $($rpool.Name) resource pool?"
		$totalvms = $rpool.ExtensionData.Vm.count
		[int]$rpshares = $pervmshares * $totalvms
		Write-Host -ForegroundColor Green -BackgroundColor Black $rpool.name
		Write-Host "Found $totalvms in the $($rpool.name) resource pool. At $pervmshares each, this pool should be set to $rpshares shares."
		Set-ResourcePool -ResourcePool $rpool.Name -CpuSharesLevel:Custom -NumCpuShares $rpshares -MemSharesLevel:Custom -NumMemShares $rpshares -Confirm:$true | Out-Null
		}
	}

Thoughts

I hope this has helped clear some confusion around resource pools, although it’s a big chunk to swallow in one bite, and I’m sure there are a lot of other competing opinions floating out there that won’t jive with mine. I’m OK with that. One thing I would like changed is the ability to set per VM shares on the resource pool, and let the pool automatically adjust for membership values.

I'll take some hosts and clusters with a side of networking and an extra helping of datastores.

I’ve spent a lot of time in those first 3 segments in the forms of servers and blades, various Microsoft operating systems, and all tiers of storage. This has given me a healthy amount of respect and experience for what’s out there. Recently, I’ve made it a goal to focus on the fourth segment, networking, and I wanted to share a bit of my experience to obtaining Cisco’s certified network associate (CCNA) certification. The ultimate journey will most likely end with CCNP and being able to work with NX_OS (Nexus) switches, but this is a good start.

Learning Materials

While I do have an undergraduate in networking, it’s been a long time since I’ve implemented anything on a router or switch. Plus, things like RIP v1, which I spent many hours learning and committing to memory, are nearly worthless now (RIP is a very ironic name). Here are the materials I used to learn and study:

CCNA Cisco Certified Network Associate Study Guide: Exam 640-802 (Amazon): This is a very detailed book written by Todd Lammle. It goes into all sorts of the nitty gritty on what you should know for the CCNA, and includes a lot of things that you may just want to know or be curious about outside of the exam.

CCENT/CCNA ICND1 640-822 Official Cert Guide (3rd Edition) (Amazon): While a bit dry and filled with Flinstones based examples, the infamous Wendell Odom does a great job and explaining things in a very simple set of terms. This thing is a tome of knowledge with some great examples and walk-throughs.

CCNA Exam Cram (Exam 640-802) (3rd Edition) (Amazon): This was actually my favorite resource because it’s written in a very conversational tone and doesn’t wander deep into the weeds at every turn. It also had a great set of practice exams (not brain dumps, I assure you) that help prepare for the style of questions on the CCNA exam.

If you want the CD that comes with any of these books, grab the dead tree edition, as it’s a bit of a pain to try and get them in Kindle format. I recommend getting the Exam Cram book in paperback, as their CD is really good. The others I did not feel added any significant value in the very limited digital content.

Home Lab

My entire home lab was simply a virtual machine running Windows 7 and GNS3. This allowed me to set up a rather complex topology of routers and switches to use while following along with the learning materials. For more details on working with GNS3, head over to their tutorial page.

The other possibility would be to purchase equipment. There’s a lot of places out there that sell CCNA/CCNP hardware kits for a couple clams, but I found GNS3 to be “good enough” for my intentions. It will not suffice for the CCNP as GNS3 cannot emulate a Catalyst switch.

Thoughts

I’ve taken exams from Microsoft, VMware, and NetApp, and to date Cisco wins the “wow this is hard” contest hands down. The exam requires 800/1000 (one required 825/1000) to pass and consists of many simulated environments where there is a terminal configured to accept line commands – you are expected to perform live commands to achieve goals set by the question (very cool!). I really enjoyed the exam content and difficulty, and I now have more respect for anyone else who has gone through the process.

I took both the ICND1 and ICND2 as opposed to the straight CCNA exam – my reason for this was to avoid taking the entire thing all over again if I did not pass, and to narrow down the content for each exam.

I personally have a preference in favor of disposable disks, as the “junk” data is deleted when a VM is powered off. It seems like a no hassle way to keep the desktops clean. And, for those wondering, the disposable and internal disks follow the OS disk in regards to datastore selection.

Details

I was working on a BladeSystem (C7000) populated with ProLiant BL460c G7 blades along with HP Virtual Connect.The HP Virtual Connect Ethernet networks were configured to tunnel VLAN traffic (pass the VLAN tags along without stripping them off) so that the distributed switches in vSphere could handle the VLAN tags. Installation of ESXi 5 on the blades went just fine. I used the “test network” option in the DCUI successfully: the blades pinged their DNS servers and resolved their DNS name without a hiccup. However, I could not attach them to vCenter, nor could I use the vSphere Client to connect to them directly.

The error I received was:

The server ‘name.of.server’ could not interpret the client’s request (the remote server returned an error: (404) not found)
Error stack:
Call “ServiceInstance.RetrieveContent” for object “ServiceInstance” on Server “address.of.server” failed.

Googling this error brings about a lot of incorrect resolutions, such as adding https to the front of the name of the server, or ensuring that the resolv.conf file contained valid DNS entires. It was obvious that DNS was working due to the fact that I could indeed ping by using FQDN values on the console of the blade.

At this point I was a little stumped. I tried a different approach. Using ethtool in the DCUI, I grabbed the firmware and driver versions and began looking around for KB articles discussing any bugs. The shipping driver version is 4.0.88.0. I managed to find this Advisory published by HP and updated on January 19th (yesterday).

Advisory: (Revision) HP BladeSystem – Emulex be2net Inbox Driver Version 4.0.88.0 Does Not Support Flex-10 or Flex Fabric Adapters on VMware ESXi 5.0 in a Virtual Connect Environment

It specifically calls out the issue:

If a VLAN tag is added to a network (for example, from the VM Portgroup, the network in VC, the external switch, etc.), then the vSphere client cannot connect to the host.

Ah ha! That was exactly the case.

Resolution

The resolution is pretty straight forward – update to the latest drivers.

The async be2net driver version 4.0.355.1 is available on the VMware website at the following URL:

http://downloads.vmware.com/d/details/dt_esxi50_emulex_be2net_403551/dHRAYnQqQHBiZHAlJQ

My method was to do EST (External Switch Tagging) on the Virtual Connect side, get the hosts connected in to vCenter, and push out the updates using VMware Update Manager (VUM). Once updated, the blades could then again use VST (Virtual Switch Tagging) without any issues. Because vSphere 5 has multi-host remediation with VUM, it doesn’t take long to patch a bunch of blades, as I set it up to just do all of them simultaneously.

I hope this helps anyone else Google searching for the answer!

Download PDF

Using the vSphere Client

Probably the most known, and easiest method, to update VMware Tools is via the vSphere Client. It’s simple and straightforward.

1. Right click on a VM in the vSphere Client
2. Choose Guest
3. Choose Install/Upgrade VMware Tools

You will be prompted to choose how you would like the upgrade to take place, either Interactively or Automatically. Along with the Automatic option comes the ability to enter some arguments, listed as “advanced options” in the GUI, that will be passed to the install.

I did a lot of web searching but could not find anyone with the updated string to put in this box after the VI3 days. Also, the official VMware documentation seems to ignore any recommended values for the advanced options box. A former colleague of mine asked what to put in the box, so I did some digging with procmon to see exactly what the advanced options box does.

Behold! The string that I have found to work is as follows:

/s /v/qn ADDLOCAL=ALL REBOOT=ReallySuppress

I’ve tested this on ESXi 4.1 and ESXi 5.0 with the same results – the VM is updated without a reboot. I will caution that you shouldn’t fully take my word for it, as perhaps some newer or older version of tools will not respond to this exact phrasing as planned. Test it out first to make sure it will go as planned. I won’t take responsibility when your production SQL box reboots.

Using a PowerCLI Script

The second method is a little more difficult, but not by much. It requires that you have installed PowerCLI onto your workstation or server.

It also gives me a chance to use my incredibly awesome PowerCLI graphic featuring Spongebob.

You can read the documentation on the cmdlet here, or you can see some further examples on VMware’s KB site or on ICT-FREAK’s page of PowerCLI one-liners, but the basic code is:

Get-VM "Name Of VM" | Update-Tools -NoReboot

You can get as fancy as you want with the Get-VM portion by using the -Location argument. In this example, I’ve nested 3 Get cmdlets to specify a datacenter, cluster, and folder to search. The script will update every VM it finds in that location.

Get-VM -Location (Get-Datacenter Foo | Get-Cluster Bar | Get-Folder Blah) | Update-Tools -NoReboot -RunAsync

Note that in this case I’ve added the -RunAsync argument to avoid having to wait for each VM’s tools to start updating. Also, it’s important to know that the Update-Tools command does not accept -WhatIf or -Confirm.

Using a Remote Execute Program (PSEXEC)

The final method is by calling the installer using some remote execution software. This could be any of your third party installers, and I use the example of PSEXEC because it’s so well known. You can use the same arguments I supplied in the vSphere Client example to accomplish this.