Archive for August, 2009

In vSphere4 there’s a new method enabling the “Distributed Power Management” options for a DPM enabled cluster. In the past you were dependent on the “wake-on-LAN” and so-called “magic packets” to suspend and resumme an ESX host. In ESX4, you able to configure a user on the iLO/IMPI/BMC/DRAC/RAC board and with the right settings get the ESX host to suspend and resumme itself via the ACPI functionality which has been around for some years.

You can configure it in the GUI like so:

That’s all well and good I guess. Follower of this blog will know I’ve recently been trying to port all my usually setup routines (commonly in my kickstart install scripts) to PowerShell. This is part of my preparation for getting ready for an ESXi only world where the days of esxcfg-commands will be long gone.

So how do you this in PowerShell. Well, I was stumped for days on this even though I was VERY close. In fact to this day, I really don’t see how my bad PS scripts didn’t work, and good one does work. I’ll just blame that on the ghost in my machine! In the end I was helped out by Stuart Radnidge of vInternals. I’m very grateful for Stu for sharing his PS script for DPM on the VMware Forums, after he responded to my forum post/question. That’s how first started delving into the more murker depths of the VMware SDK APIs. Stu has a very nice post explaining (for idiots/mere mortals like me) how deep into the SDK using PowerShell has your pick-axe. It’s called PowerShell Mastery - http://vinternals.com/powercli-mastery/.

Additionally, if you into DPM the other thing Stu has shown, is how to create DPM user account on a HP ILO using the HP ILO command-line utilities. That means the entire job (create DPM using in ILO and configure on ESX host) can be scripted. Of course the tricky thing to script I would imagine is getting hold of the MAC address of ILO. I imagine you can do that with the HP ILO CLI commands…

http://vinternals.com/2009/08/create-users-with-the-hp-ilo-cli/

Anyway, that’s the background. Here’s Stu’s DPM script. Stu wasn’t bothered about me giving him all credit –  but I think he deserves it. All he requested was some gratious swearing in this blog post. (Well, thank fuck for that, I thought he was gonna ask for book royalities!)

I’ve slight modified it (no major surgery) so it matches the conventions used in other samples on RTFM…

$vmhost = “esx4.vi4book.com”
$login = “vmware_dpm_user”
$password = “password”
$hostview = get-vmhost $vmhost | % {Get-View $_.Id}
$IpmiInfo = New-Object Vmware.Vim.HostIpmiInfo
$IpmiInfo.BmcIpAddress = “192.168.3.204″
$IpmiInfo.BmcMacAddress = “00:16:35:37:F8:02″
$IpmiInfo.Login = $login
$IpmiInfo.Password = $password
$hostview.UpdateIpmi($IpmiInfo)

Update:

Since writing this post I think I’ve figured out why Stu’s script worked and mine didn’t. He was including | % {Get-View $_.Id} and I wasn’t (not sure if that was significant). Additionally, it seems to be the case that the ESX host must added into vCenter first for the script to work. I found DPM didn’t have to be enabled on the cluster to allow the DPM configuration to go through.

Next years VMworld will be:

October 11 – 14 at the Bella Centre, Copenhagen.

Tell you what I’m really pleased about? That’s it’s not in France! No offense to my Galic chums but I was getting a bit tired of being in the South France for 3 years in a row, and then Paris Disney the year before that.

VMworld EU, follows hard upon VMworld US which again will be held on August 30 – September 2 at the Moscone Center, San Francisco.

It will be tricky for us EU people to decide which event to attend. The US one to keep in touch with all our US contacts. The EU one because it will be more cost affective and less jetlag OR both?!?!?

How did I find this out. From a retweet from: http://twitter.com/vnagornyi

Who read the “agenda at-a-glance” which is available as PDF on VMware’s website. Last page…

http://www.vmworld.com/servlet/JiveServlet/downloadBody/3707-102-1-4759/vmworld2009_program-guide.pdf

I’ve been watching the MS blog for sometime – and noticed there had been a couple of articles about the importance of having a small footprint. I’d been mean to read those post fully an answer their claims. Anyway, the folks at VMware have taken it upon themselves to rebut these claims here.

http://blogs.vmware.com/virtualreality/2009/08/our-position-on-hypervisor-footprints-patching-vulnerabilities-and-whatever-else-microsoft-wants-to-throw-into-a-blog-post.html

But if you really can’t be bothered to do so all you need to see is this image. ESXi is that very small circle in blue. Yeah, its so small I didn’t see it at first glace either….

This is a funny video!

This video shows you how install the Rapid Clone Utility, and use it to create virtual desktops very quickly – and also how to turn on deduplication and grow a NetApp volume.

Some thoughts on the RCU from making the video (it took a couple of time get video working right, due to user errors such as not having the right key for Windows XP!). Firstly, I can spot to very small tweaks that could made with the RCU. It doesn’t allow me to control where the VMs are created from a folder or resource pool perspective. All the new virtual desktops are created in the same folder as the source template – and there’s no ability to place the rapid cloned VMs into a resource pool – and this can lead to big long flat list of VMs – which ain’t pretty. Especially, in a VDI environment where there could be hundreds of virtual desktops. You might think this is a minor issue, after all you can drag & drop the VMs after the cloning process to the right VM folder/resource pool. Erm, well not quite. During the deploy process you can ask the RCU to register the new virtual desktops with VMware View/XenDesktop brokers. I dunno about XenDesktop but currently View is very happy about you moving vCenter objects around. If you don’t know what I mean check out this recent KB article that Chris Wolf brought to my attention:

http://www.rtfm-ed.co.uk/?p=1463

If you watch the video you will see the RCU automagically create a NFS volume/mount point for me, and present it to my ESX host (if feels a lot like the way SRM automagically presents replicated LUNs during a test of DR plan). The way it does this by modifying the permissions on a newly create volume/mount point – adding the ESX hosts in by IP address. What IP address? Well, every single VMKernel port on the ESX host – so what you see – is not just IP address used to speak to the filer (172.168.3.0 in my case) but also VMotion ports (10.0.0.0 in my case) and if you using ESX4i, the management ports also get listed (192.168.3.0 in my case) because they are “vmkernel ports” too. I’m perhaps griping a little here. Because firstly, its much faster/automated to use the RCU to create/permission up the volume/mount point – and secondly I don’t think the vNetwork/vStorage allow for the storage vendors like NetApp and EMC to work out – that this little pig goes to VMotion, and that this little pig goes to management, and this little piggy goes wee-wee all the way to NetApp FSA or EMC Celerra… Anyway, it wasn’t too much of chore to highlight the IP address that shouldn’t be there, and remove them – compared to the core having to do all the work manually (including mounting the NAS/NFS mount point to each and every ESX host!!!)

The RCU cloning process is pretty good, but it reliant on the “Guest Customization” settings and Sysprep process. If you know (and loathe) sysprep well, you will know its not especially quick – and also it takes some work with sysprep.inf files to make sure the computer account objects of the virtual desktops are dropped in the right OU. That’s important if your unlucky enough to have to use MS GPOs to control the desktop environment. In contrast VMware’s View Server now ships with a linked-clone feature and something called “QuickPrep”. QuickPrep creates computer account objects for the virtual desktops – and ALSO puts them into the right OU for your GPOs. What I would love is a combo of both – RCU with Linked-Clones would be quiet cool…

Another interesting thing about the RCU is the “destroy” feature – powers of all the VMs, unregisters/deletes the VMs, and then goes off and destroys the volume/mount point. Clearly, despite all the warnings anyone using the RCU will have “click: Engage Brain” before they use this. Currently, what the destroy option DOES NOT do, is delist the virtual desktop pool from the View Broker. I guess that’s OK, after sometimes you will want to keep that metadata, and other times you won’t. The only trouble is you will have to remember to un-entitle (is that word?) the virtual desktop pool because until you did it would be listed and viewable by end-users – with no virtual desktop behind them. That would make you popular “yeah, we know you can see it on the list… but we destroyed them all this morning… sorry, about that… have nice day… bye…”

All of this is all well and good. But there’s a feature of RCU that makes me smile – and it’s kinda tucked away. That’s the ability to just right click a NAS/NFS volume and make it bigger. You just right click and type in the new size, and jobs done. Neat…

Anyway, there’s been a lot of traction on the blogs this week, along the lines of “which storage vendors has the best integration tools”. I’m keeping well out of the debate because I don’t wanna caught between two vendors who I have enormous respect for (EMC and NetApp). But my plan is to do a session with NetApp on their integration tools – and do videos of them. AND do the exact same with EMC. I will blog about both not to do some lame compare/contrast – but just to showcase how both vendors are working hard to integrate their array functionality into the new vSphere vStorage APIs. There’s a lot marketing guff around the new APIs and wonderous they are (and they are) but they really don’t come alive until you see them on screen and know how what their advantages and disadvantages are.

If you want to see the video click on…

NetApp RCU [ 0:01 ] Play Now | Play in Popup | Download

About a week or so go I did a webex with vWire’s Mike Connors – in preparation to evaluating their product. This has been on my desk for months, and I finally got round to the product this week.

Two things really triggered my evaluation – firstly, the product has been recently released and is now compatible with vSphere4. I’ve been using vSphere4 since the beta programme, and all my lab kit is dedicated to run it, with no spare capcity for run (ahem) “legacy” versions of ESX and vCenter. Secondly, over the last couple of days/weeks/month I’ve been immersing myself in PowerShell – and vWire has a PowerShell component to it which is used to re-mediate problems it discovers.

So anyway, what is vWire? On the front side its a monitoring/health-check system that doesn’t just look at your vCenter DB, but also uses the log files of ESX and the SDK to build up a picture of your environment. So its very “operational” focused in managing a pre-configured environment.  I think this maybe its main selling point – that data it collects and reports with is denser than what the core vSphere4 product does for you. In this respect its simplifying the process of collating and acting on all that log info which is held by VMware in many different places. Additionally it has an alarm/alert function – together with a remediation engine – which allows PowerShell scripts to go out and fix the problem.

What’s interesting about that – is whilst it all well and good to fix problems with PowerShell – what’s missing is a engine to trigger the remediation process. Whilst vCenter does now ship with lots of conditions & alarms – an ability to trigger scripts. Alarm/Alert system in vCenter is not extensible by the average admin unless you fancy writing your own application & plug-ins for vCenter.

As a tease to evaluation the vWire product there are some free tools on the site including OpsCheck and ConfigCheck. OpsCheck confirms if VMotion is correctly configured in your environment, whereas ConfigCheck confirms the security status/configuration of your ESX Host.

http://www.vwire.com/free-tools/

Additionally, vWire has “Community Content Library” where vWire uses can upload and share their solutions which they have created within vWire. These can be download by other users to extend the core functionality of vWire. It’s perhaps fair to say that vWire has some work to build and extend this community – as most of the contributions come from vWire staff such as now legendary, Steve Beaver.

Well, that’s the overview of the product – now for my evaluation…

(more…)

My pal and fellow Blogger, David Davis new VMware vSphere video training course was released yesterday. It is 17 hours of video training that includes multiple formats like AVI, WMV, iPod/iPhone, and MP3. It starts from the planning and implementation of vSphere and moves all the way into advanced features like FT, Data Recovery, and vDS. The URL for the course is at:

http://www.trainsignal.com/VMware-vSphere-Training-P76.aspx

Dave will keep producing and offering more videos to those who purchase the course and the course will soon be all online, streaming over the Internet.

On the VMWorld front, Train Signal is sponsoring the VMworld 2009 Sunday night extravaganza -

http://www.vmunderground.com/vmworld-2009-sunday-night-extravaganza

Additionally, Train Signal is sponsoring the VMworld 2009 Run -

http://www.active.com/page/Event_Details.htm?event_id=1786485

Finally, Dave will be doing a “signing” / meet the author session on Wednesday at the VMworld bookstore from 12-1

Alongside EMC, NetApp has very generously come good on a promise of storage for my lab environment. The main reason for this is for the new SRM book I’m writing. Well, I say writing I haven’t actually started work on the new edition yet. I felt I needed some downtime from being an author – so work begins on the all new SRM book after VMworld.

This time around I’ve informally asked some folks in the community who I respect to contribute the expertise to the project – so far I’ve approached Adam Carter (Lefthand Networks), Chad Sakac (EMC) and Vaughn Stewart (NetApp) – along side that I asked Scott Lowe if he would be willing to contribute because I know he is very strong on the NetApp side of things, and also a respect author like me.

The whole idea is to try to make the book a lot more “vendor” neutral than the first edition. The previous edition was quite heavily dominated by the Lefthand Network’s VSA, which largely resulted with LHN coming forward with NFRs for their VSA appliance.

My other plan is to make the next edition for the SRM book free. That it will be a free PDF to download from LULU, or you will be able to buy (at cost) a printed copy of the book. Either that, or a nominal $5/10 buck mark-up that will either go to charity or used to get the contributors nice and drunk at the next VMworld we all attend. How’s that for encouragement.

Anyway, once again – thanks NetApp for the donation. Now, if I could only persuade HP, IBM or Dell to replace my ageing Proliants, with servers that support VMware FT, I’d be as happy as a sandman. That’s one hell of a hint, eh?

This one has been knocking about in my head from sometime. One of the main manual configurations I have to do with ESXi is settings a number of IP related parameters:

• Hostname
• IP
• Subnet Mask
• Default Gateway
• Primary & Secondary DNS Addresses
• DNS Suffix

Before I can run my PowersHell configuration scripts against the ESX it has to have an IP address and hostname that can be resolved, otherwise my PowersHell scripts go somewhat doolally. So for sometime I have been toying with despensing with a static configuration, and using DHCP to configure the ESXi host.

Horrifying isn’t it? If your from EMEA like me, the idea of giving an physical server a DHCP address is unspoken tabboo. You just DON’T create that kind of dependency between the physical server and the outside world. Interestingly, I stuck rigidly to this view until last year, when I was stopped in my tracks by PowerPoint slide which said “use DHCP to solve this problem”. I was aghast. You’ll never get that past anyone in EMEA, I said – some of them can’t even get approval for DHCP for PXE booting in a server room LAN. And… I guess that for many that will remain the case. It seems this culture is perhaps different across the pond…

In a very informal poll of my US chums – the issue of using DHCP to configure ESXi seemed to garner the response – if it that’s works for you, what’s the problem – I do it here… Perhaps all these years I’ve got stuck in a rut with DHCP?

So today, I thought I would experiment with reconfiguring the Linux DHCP service that runs in my UDA in my lab environment to lease a static “client-reservation” to my ESX host. I did that by adding:

host esx4 {
option host-name “esx4″;
option domain-name “vi4book.com”;
option routers 192.168.3.130, 192.168.3.199;
option domain-name-servers 192.168.3.130, 192.168.3.199;
hardware ethernet 00:15:60:AC:E4:40;
fixed-address 192.168.3.104;
}

I got the MAC address of the ethernet card from the HP ILO. Then I did a factor reset of the ESXi host, which left me with just the root password to setup on boot-up. Everything now is done by combination of DHCP with PowersHell….

Well, the licensing part of this took me the better part of day to work out. That’s because I’m more or less crap at decoding the API SDK referrence guide:

http://www.vmware.com/support/developer/vc-sdk/visdk400pubs/ReferenceGuide/index.html

My favourite part of the license script is the $licassman variable. It could be used to describe a number of people I’ve worked with…

I want learned today is the add-host cmd-let has new parameter it didn’t have in the beta/RC1 of vSphere4. The -force switch is used to accept the default/untrusted SSL SHA certificate that comes from clean installation/first boot of an ESX host.

$vmhost = “esx4.vi4book.com”

add-vmhost $vmhost -location (get-datacenter -name ‘SanFran DataCenter’ | get-folder -name ‘AMD Hosts’ | get-cluster -name ‘AMD Cluster 2′) -user root -password password -force: $true

$targethostMoRef = (get-VMHost $vmhost  | get-view).MoRef
$si = Get-View ServiceInstance
$LicManRef=$si.Content.LicenseManager
$LicManView=Get-View $LicManRef
$licassman = Get-View $LicManView.LicenseAssignmentManager
$licassman.UpdateAssignedLicense($targethostMoRef.value,”YOUR LIC KEY”,”vSphere4 Enterprise Plus (1-12 cores per CPU”)