2009-04-27

Will we be Forced to Pay More – pt. 2

To continue my previous post about the Enterprise license not being available after December 15, 2009. I received several comments – some on the blog and some I received offline, but the picture has become clear.

When vSphere launches, there will be 4 tiers

  • Standard
  • Advanced
  • Enterprise
  • Enterprise Plus

Thereafter:

  1. The Foundation License will be will be terminated immediately
  2. Enterprise will stay available until December 15, 2009.

Starting December 16th, 2009 there will be 3 tiers only

  • Standard
  • Advanced
  • Enterprise Plus

Below are a few slides that confirm the fact.

image

image

image

So as you can see from the above, starting from December 16th, 2009 you will not be able to purchase vSphere Enterprise only Enterprise Plus.

Why did I label this post “Will we be Forced to pay more?”. I mean force is a pretty harsh word, but unfortunately, VMware will be leaving us no choice.

Let me explain why. Any Enterprise Customer makes extensive use of DRS and will find a huge benefit in using DPM. Those features come in Enterprise and Enterprise Plus only.

Take this hypothetical scenario for example. I have a 3 Clusters running with 5 ESX 3.5  Enterprise Licensed Hosts (with 2 sockets each). vSphere launches and with the launch I automatically receive with the new version the additional features of vShield Zones, Thin Provisioning, FT, Hot-Add and Data Recovery (don’t take me the wrong way – this is a lot and all highly useful and we should be extremely grateful). But for all different kinds of reasons we decide not to upgrade our ESX host licenses to Enterprise Plus. (I mean for that size environment Upgrades costs will be $295 * 2 * 15 = $8,850).

Comes January 1st, 2010. We need to add an additional host to each cluster. Now we need the Enterprise features (like all our other hosts). We cannot purchase an Enterprise license only Enterprise Plus.

  1. It costs more approximately $750 more per socket.
  2. But we do get the additional features of Host Profiles and Distributed Switch – so we should be happy shouldn’t we?

Here is the catch! Point 2 – is not completely true. Your will get the license to enable Host Profiles and Distributed Switches – but you will not be able to use it. And why not? Because in order for these features to work they have be applied cluster-wide to all hosts in the cluster. So in this case, we came out on the short end of the stick on both counts.

We would have been better off with doing the upgrade when we had the opportunity.

My thoughts – initially not pleased. But I will have some more to say about that in a later post.

And you?

2009-04-23

Another “Myth Buster” - Yeah Right!

Extra Layers – cost more – Bah Rubbish !! It is the Enterprise features that Hyper-V (even in R2 – that to remind you all is at least 6 months away) does not have - that makes it all worthwhile!

Over Commit – don’t nag about something you cannot do. How many times do we have say this? This is a feature that we use. All the time and enjoy using it – and we get better results than you bring about in your video.

You are going to get bashed again with this one… Come on Blogosphere. Bring it on!!!!!!


Microsoft Virtualization - Virtualization without Taxation

Will we be Forced to Pay More?

VMware announce their licensing changes and I wrote about it in a previous post. I noticed something on a document that (I gather) was released by VMware Marketing and brought to the blogosphere by Jason Boche.

The title is, “VMware vSphere 4 Cheat Sheet”.

If you look carefully at the above document you will see on on page 3 of the document it is stated that vSphere Enterprise version is going to be available for sale only until Dec. 15, 2009. below is the screenshot.

 

This was noted again on page 4 as well.

This does not appear on the same graphic which is in the vSphere Pricing, Packaging, and Licensing Whitepaper on page 7.

I truly and sincerely hope that this was a mistake. I mean if Enterprise Edition will only be available for the interim period until the end of the year, and then all new purchases will have to be Enterprise Plus licenses starting after December 15th, 2009, then I am sorry VMware - but you are going to have a lot of extremely unhappy customers (to put it mildly).

It would be nice to have someone from VMware to confirm if this is a mistake and amend the document. And if not to explain themselves.

2009-04-22

Say Hello to vSphere PowerCLI

So all the buzz have quieted down – I think have approx 600% less traffic coming in through Twitter than I had yesterday.

All the news about the products names – version changes etc. etc. But I do not recall that there was a mention of one small but extremely important part that will also undergo changes.

The SDK’s and Toolkits.

This I got from the VMware Developer Center Blog. What is new in vSphere SDKs &Toolkits....

SDK

Say goodbye to the to the old VI Toolkit and the RCLI and welcome the new names for our old and trusted tools

2009-04-21

vSphere 4 – So What is new?

Well I think my speculation was not far off. if you read my previous post I speculated that we would not get anything 

Unofficial Release Date - May 21, 2009. We got this from the day that you can start purchasing new licenses

Jason Boche posted a few noticeable items about the licensing changes.

Let me see I can make a bit of sense for what we are getting. All of this comes from the
VMware vSphere Pricing Packaging and Licensing Overview


Versions

Some of the version names have changed (I mean what would we not live without another name change).Below is a comparison between the two:

versions

Are we getting anything new with the new names – yes we are!

features

Pricing:

The biggest change here is that licenses are per physical Socket and not licensed per two sockets as in the past (and is it only or did the prices just go up??)

Prices

 

Licensing:

Licensing

Now besides that all Twitter is going nuts, all the bloggers are going nuts and when the dust settles and things cool down I promise some more detail and insight later on down the road...

2009-04-12

Is P2V always the best solution?

I saw this post post on Simon Gallagher's Blog Using Virtualization to Extend The Hardware Lifecycle. Simon made a good point. Can we take the physical hardware out of the equation? Well I guess that is exactly what VMware and virtualization is doing. We are no longer dependant on physical servers any more.

I would like to however raise another point of view. Correct, the easiest solution for a system residing on aging hardware is to P2V the system to a VM – and it seems that I have solved my problem. But have I? I posted a question about 9 months ago with this exact same thought.

Do you really want to carry over legacy Operating Systems and applications and continue to support them forever?

You might find yourself with legacy operating systems, applications that Vendors no longer support, personnel do not have enough knowledge of these old operating systems, and will not know how to solve problems when they arise.

Classic example: A Windows NT4 Server that has a legacy application was installed on it. The product was developed and installed over 6 years ago. A new system was requisitioned and designed to replace the old, for obvious reasons. The original company that developed the product no longer exists, staff that understand and support the application are no longer employed by the company. A project was planned and executed over the period of a year to replace the old system. The project was completed on schedule with close to 100% success. The new application Owner agreed that the old system would stay available for a period of maximum 6 months after going live with the new system and the old data would not be required to be migrated to the new system. Those six months have come and passed. When push came to shove – and we wanted to power off the server – the application owner (and thereafter it became a management decision) demanded that the old server stay online.

The hardware was over 7 years old. No parts, no service, and no guarantee that it could be restored successfully if the hardware was to die. The only solution that we were able to come to was to P2V the machine to a VM. Of course the process works, but now the organization is stuck with a NT4 server that has to be supported. Operating System upgrade to Server 2003 is being planned and tested.

I do sincerely suggest that you create a policy that a VM/Server should have its lifespan defined according to your company’s requirements. Any VM/Server that has passed that lifespan should not be automatically migrated to a VM to remove the hardware dependency – but should be installed as a new OS and have the applications on the old VM/Server migrated to the new OS. Just the same as if your hardware would come of age and have to be replaced.

Do not blindly P2V systems, just because you can.

You will regret doing so in the long run.

I for one do not want ancient Windows/Linux systems in my datacenter. Would you?

2009-04-07

April 21st – What will we get?

So many have already posted about the Virtualization announcement that is planned for April 21st, 2009 (here, here, here, here, here, here, here, here, here, here). You can see from the amount of announcements on the Blogosphere that this is a highly anticipated event. (By the way all of the blogs posted above have a huge wealth of information that I highly recommend you frequent whenever you have a chance)

So I put up a balloon on Twitter this morning with the question, “Is this going to be the launch of vSphere (a.k.a. ESX 4) or will this be the announcement of when the launch date will be?”

The only answer I actually got from the Twittersphere was that is is an announcement – what it will contain is not yet disclosed (which – in VMware’s defense – I must say would sort of “be missing the point” if they removed the suspense. The hype around this announcement is HUGE!! I mean I only gave 12 examples above of bloggers that posted about it in the past 24 hours – I am sure there are at least double the amount of posts that I might have missed.

What will we get? What will the announcement be? the way I see it - three different alternatives.

  • VMware will release their new version of vSphere. It has been around in Beta since September 2008 (I have personally filled out countless amounts of surveys and feedback forms with my input on all aspects of the product and all the additional functionality that will be released with it).
  • VMware will announce the release date for vSphere.
  • Something completely new will come out – that has not been circumventing around the virtualization world.

Now I have some thoughts on each of the above.

Option One. vSphere has been in Private Beta since September 2008 (6 months in total). I was one of a number (exactly how many I do not know) of individuals that were given access to the new version – even before it was called vSphere. Testing, installing, playing around has been a good learning experience, has helped the VMware development team to make the product better and add in improvements across the board. Besides all the public information that has been presented, in VMworld 2009 in Cannes, by bloggers which have rightfully or not disclosed features / screenshots of the upcoming product, and has taught me a lot about testing and learning a new product.

But, and for me it is a big BUT.. I do not want to receive a product that is not complete yet. We all remember the fiasco that happened with VMware and the time bomb bug last August, which I think most of us agreed that the reason this happened was because of the rush to get out a product and Microsoft’s Hyper-V in the works. Of course mistakes happen, but I am wondering, if the product is to be released in two weeks from today – is it ready? Are we in for more some more bugs and embarrassments due to cutting corners to meet deadlines? Believe you me – I know that no software release is perfect, bug-free. There is always a trade off between perfectionism and time. Someone once told me that when developing software you have to draw the line where the delivery of the product is not perfect but where the delivery is good enough to go out but financially tangible and worthwhile. But I personally think that this 6 month period of broader testing is not enough. The amount of additional features, improvements new technologies, practically a complete overhaul - will be a big incentive to persuade VI Admins worldwide to move to the new version. But as always – without extensive testing, and waiting for what I would call a “let the bugs emerge” period, the new version will have to wait a while until I deploy it.

Now that is out of the way – let’s go to Option Two. I will be happy to finally have a release date – which will allow for proper planning for the move to the next version, adjust licensing costs – new hardware requisitions, training and career paths. VMware partners will also have time to finish their new products and technologies that will compliment and improve the next version. I do not know many people who like live without knowing what is in store for them in the future

If this is the case – the HUGE HYPE – will not be warranted. I mean what else can they tell us that has not already been announced on numerous occasions in the past 6 months?

Option Three. This one is intriguing.. What has been going on under the covers? I mean Chad Sakac has been hinting on this on numerous occasions – for example. I am sure the other partners have their secrets as well. Could there be something that will surprise us all? I don’t think so – a secret like that is really, really hard to keep. We still have two weeks till the announcement so maybe we will learn some more as the date gets closer.

Personally I place my bets on option 2 with something additional from option 3 – but nothing earth shattering.

What are your speculations?

VI Client vulnerability - Security Alert

I picked this one up on Twitter and Secunia. Another reason to to upgrade to VC 2.5 U4

VMware Virtual Infrastructure Client Password Disclosure Weakness

Secunia Advisory:
SA34585

Release Date:
2009-04-06

Critical:

Not critical

Impact:
Exposure of sensitive information

Where:
Local system

Solution Status:
Vendor Patch

Description:
A weakness has been reported in VMware Virtual Infrastructure Client, which can be exploited by malicious, local users to disclose sensitive information.
The weakness is caused due to the application improperly storing passwords in memory. This can be exploited to potentially disclose the VirtualCenter Server password by reading the memory of the client process.

Solution:
Update VirtualCenter, ESX, or ESXi to a fixed version, and install the included client.
VMware VirtualCenter:
Update to version 2.5 Update 4.
ESXi 3.5:
Apply patch ESXe350-200903201-O-UG (also included in ESXi 3.5 Update 4).
ESX 3.5:
Apply patch ESX350-200903201-UG (also included in ESX 3.5 Update 4).
Provided and/or discovered by:
The vendor credits Craig Marshall.
Original Advisory:
VMSA-2009-0005:
http://lists.vmware.com/pipermail/security-announce/2009/000054.html

2009-04-02

New VMware Product Releases

NO! it is not the new ESX version! (there are other products that VMware sells besides ESX you know)

Besides Update 4 for ESX and ESX3i that were released last week:

VMware Player 2.5.2

VMware Workstation 6.5.2

VMware Server 2.0.1

What's New

With this release of VMware Player, certain new features and support have been added.

Support for New Guest Operating Systems
VMware provides support for the following operating systems for Player 2.5.2:
  • Windows Vista Service Pack 1 and Service Pack 2
  • Asianux Server 3.0 Service Pack 1
  • openSUSE 11.1
  • Ubuntu 8.10
  • Ubuntu 8.04 LTS
VMware provides experimental support to the following operating systems for Player:
  • Asianux Server 3 Service Pack 1
  • Fedora 11
  • FreeBSD 7.1
  • Mandriva Linux 2009
  • Microsoft Windows 7
  • Novell SLE11.0
  • Red Hat Enterprise Linux 5.3
  • Red Hat Enterprise Linux 4.8
  • Sun Solaris 10 Update 6
  • Ubuntu 9.04

Besides the new operating systems that were supported – there are several other issues that were addressed with each release.

Go forth and Virtualize!!

2009-04-01

Not another esxtop post ...

Well Actually yes. Because there is so much useful information about this reporting and troubleshooting tool, I wanted to add my own.

I was pointed to very useful document regarding esxtop. Granted it has been a relatively popular topic lately, Both Jason and Duncan have very good posts on the subject

So take a look at the document Interpreting esxtop Statistics

vCenter 2.5 U4 – My take….

Well I have been meaning to put up my experiences out on the web about what has happened with my attempts (and partial success) with the installation of vCenter 2.5 U4.

Now for me the nicest new additions in this update were:

  • Guest Operating System Customization Improvements
  • Performance Overview Charts

So I set out to install this update. Pretty straight forward. The install went smoothly – Database was upgraded – Update manager the whole works.

Then came the Performance overview plug-in. Instructions were pretty simple.

All seems to go ok – but then I ran into the problem which this KB article solved for me.

So now I have the plug-in working correctly – or at least I thought I did.

I was not getting some “chart cannot be displayed” errors on some of my hosts. (see below)

Working

Not Working

 

Some clusters would work – others would not. Some hosts in a cluster would work and others would not.

Now after some debugging of the issue I found:

  • The issue was consistent on the same hosts all the time.
  • No matter what I did (and tried all the relevant KB’s – and even re-installing vCenter from scratch – using the same DB) would not help.
  • In the stats.log from the vCenter server I was seeing:

         

      When querying (via GUI) a working host the stats.log returns:

      ----

      [30 Mar 09, 15:35:31] [INFO] com.vmware.vim.stats.webui.action.ViClientRequestAction : Request: :8086/statsreport/vicr.do">:8086/statsreport/vicr.do">http://<IPADDR>:8086/statsreport/vicr.do ? &moref=HostSystem:host-11041&sessionId=9A123CFD8AA-058F-4068-AC7C-66C66C2BDC7F&locale=en_US&serviceUrl=https://<hostname>/sdk

      [30 Mar 09, 15:35:31] [INFO] com.vmware.vim.stats.webui.action.ViClientRequestAction : User (vmware_soap_session="AFCEC24C-1B37-414D-92E0-3523AC1BDED3F8") successfully authorized...

      [30 Mar 09, 15:35:31] [INFO] com.vmware.vim.stats.webui.action.ViClientRequestAction : Redirect to corresponig MO Action (HostSystem)...

      [30 Mar 09, 15:35:31] [WARN] com.vmware.vim.stats.webui.util.JspHelper : The following metrics does not exist ini the VC: (swapinRate,average,3)

      [30 Mar 09, 15:35:31] [WARN] com.vmware.vim.stats.webui.util.JspHelper : The following report will no be displayed because one of its metric s is not present in the VC:hostMemRatesReport [30 Mar 09, 15:35:31] [WARN] com.vmware.vim.stats.webui.util.JspHelper : The following metrics does not exist ini the VC: (maxTotalLatency,latest,3)

      [30 Mar 09, 15:35:31] [WARN] com.vmware.vim.stats.webui.util.JspHelper : The following report will no be displayed because one of its metric s is not present in the VC:hostDiskTotalLatencyReport

       

      When querying a non working host (via GUI) we get this is in the log

      ----

      [30 Mar 09, 15:34:44] [INFO] com.vmware.vim.stats.webui.action.ViClientRequestAction : Request: :8086/statsreport/vicr.do">:8086/statsreport/vicr.do">http://<ipaddr>:8086/statsreport/vicr.do ? locale=en_US&timeRange=daily&startDate=-1&pageIndex=0&moref=HostSystem%3Ahost-3589&mode=home&sessionId=9ACFD8A2A-058F-4068-AC7C-66C626C2BDC7F&endDate=-1&pageSize=2&serviceUrl=https%3A%2F%2F<HOSTNAME>%2Fsdk

      [30 Mar 09, 15:34:44] [INFO] com.vmware.vim.stats.webui.action.ViClientRequestAction : User (vmware_soap_session="AFCEC24C-1B37-414D-92E0-35AC1BDED3F8") successfully authorized...

      [30 Mar 09, 15:34:44] [INFO] com.vmware.vim.stats.webui.action.ViClientRequestAction : Redirect to corresponig MO Action (HostSystem)...

      [30 Mar 09, 15:34:44] [WARN] com.vmware.vim.stats.webui.util.JspHelper : The following metrics does not exist ini the VC: (swapinRate,average,3)

      [30 Mar 09, 15:34:44] [WARN] com.vmware.vim.stats.webui.util.JspHelper : The following report will no be displayed because one of its metric s is not present in the VC:hostMemRatesReport [30 Mar 09, 15:34:44] [WARN] com.vmware.vim.stats.webui.util.JspHelper : The following metrics does not exist ini the VC: (maxTotalLatency,latest,3)

      [30 Mar 09, 15:34:44] [WARN] com.vmware.vim.stats.webui.util.JspHelper : The following report will no be displayed because one of its metric s is not present in the VC:hostDiskTotalLatencyReport [30 Mar 09, 15:34:44] [ERROR] com.vmware.vim.stats.webui.chart.StatsChartServlet : Unable to generate performance chart image. Details: null

      java.lang.NullPointerException

When running a profiler on the SQL server side – I was able to catch the query (which is actually a very nice SQL query that will  give you the info that should be presented in the tab – but that is a different post..) When running the query against the SQL Database itself it was returning the data perfectly for all hosts – those that would present properly in the tab and those that would not present at all. When the tab would not display correctly on a host – then the query would not even be presented to the SQL database – which leads me to think that something screwy is going on with the JAVA or in the SQL tables here.

SR was opened with support.

The workaround that I was provided from support was to completely remove the host from the cluster and add it in again. Now this made me think – of course this would solve my problem. Because there is something not working correctly in one of the tables – and when you remove a host from the VC and add it in again, it receives a new ID – so of course this would solve the problem. New tables – new data – makes a lot of sense. But… What happens to all the historical data on the host once you remove it? You guessed right – GONE!Forever!!!

So for me that workaround is not acceptable.

The problem is a known bug (or as the Support technician put it “Boog”) that currently does not have a solution and that will be fixed in the next release of vCenter (which I suppose will be the GA of vSphere).

Now something that worried me actually – and I still have not had the time to check it out thoroughly – was when I saw the release of vCenter Mobile Access, the charts that are displayed – seem to be very similar to what is displayed on the Performance Overview tab – so I am wondering if I will get the same problems when I get to install that. Will have to wait and see.

Life ain’t easy!!