2010-12-29

ESXi has a 32mb Memory footprint

I was listening to an very interesting presentation this evening from Olivier Cremel, Principal Engineer
ESXi Internals: Better Understanding for Better Management and Troubleshooting

The ESXi Architecture document actually states that the memory footprint of the hypervisor is 32mb

image

Eric Gray, also put up a post regarding the hypervisor footprint - again 32mb

The hypervisor is loaded into memory into a RAMDISK. But where exactly can you see the size of the RAMDISK in which the hypervisor runs, and is it actually 32mb?

Two places:

1. In the System Resource Allocation settings for your ESXi host.

image

Here you can see that the size of 32MB is reserved - so that it will always be available - but it is also limited at 32MB so that it does not grow larger than the specified size.

2. In the ESXi host itself.

image

As you can see from the output below

~ # vdf --help
vdf: invalid option -- -
Gather stats about visorfs
visorfsdf [filepath]
   -h       multiple instead of blocks
   -p       resource pool view

As you can see from the output above the MAINSYS partition is the set 32MB.

Of course because all the entities above are resource pools, they can be configured to your needs - but they should only be done with specific guidance from VMware support.

A nice bit of info..

2010-12-28

What is in Store for 2011 and How to Keep up?

2010 has come and gone. It has been a great year, seriously, there are so many things I can point out that continue to amaze me, that makes me wonder sometimes what is it that I do that has  brought these great times upon me.

Two contributions of mine that were published lately were predictions on what is store for us in 2011.

Predictions for 2011 by the virtualization community - Veeam 

2011 server virtualization predictions from our advisory board - SearchServerVirtualization.com

I will not repeat any of the the predictions that have been published above - but I would like to add a small something on a personal note.

Keeping up with the technological advances is becoming harder. Anyone that does not agree - put up your hand! More and more products are being released, by all of the major players in the industry.Be it virtualization, be it the supporting Eco shell around it (monitoring, management, planning, security etc. etc.) , be it the hardware vendors, the storage vendors the sausage vendors, (hey not the sausage people - just wanted to check if you were paying attention). If you are not developing a product that will be connected with virtualization in some way - then you are losing out on a really large potential market.

If I were to dedicate all my time (which I cannot - I have a life - and a day job) to only testing all the new products, technologies that are being released, developed on a daily basis, then perhaps, maybe I would be able to keep up. Maybe. Only very few of us can actually do this there just is not enough time.

So how to keep up? I have a few suggestions.

  1. Twitter. Twitter is a unbelievable source of information, it is actually sometimes scary and can become quite an addiction, that is why you have to control how you use Twitter. If you have not already joined in I would suggest the following lists to follow.

    If you you were to follow the people in these lists (and some of them overlap) then you will most probably follow most of the people that are active in the VMware/virtualization world - that would give you a good push start.

  2. PlanetV12N. If this is the only thing you can do - then subscribe to this feed. This is a great source of information which is something that I make time (actually more than once) daily to go over the blog articles and news that was released.

    If you have any other good sources - then please feel free to share!

  3. Concentrate on certain projects and products, not everything has to be tested, evaluated. Pick and choose what you find to be relevant to your business and your needs. I suppose, that if the need arises (and is justified) for a certain product to be tested, then the time will be allocated and freed for you.

2011 is going to be a whirlwind of a year - with a great amount of technology and ideas that will emerge and evolve and go GA

I hope to get back to more regular blogging, now that I have completed a major project, that you will be hearing about in the not very distant future.

Thank you all for a great year, and here is looking forward to an even better one!!!

2010-12-23

Visio Stencils - Happy Holiday's to you all!

*****Update June 11, 2012******

The new version is available here

*****************************

This has been long overdue.
I would like to share with you the new Visio Stencils that I have completed.
But first the legal part
These are the official icons and diagrams of VMware, Inc. VMware allows the use of these icons and diagrams in derivative works by VMware Community members to illustrate virtualization concepts and IT architectures only if the VMware copyright and terms of use are clearly displayed. The VMware icons and diagrams cannot be altered in any way. VMware does not endorse or make any representations about third party information included in this document, nor does the inclusion of any VMware icon or diagram in this document imply such an endorsement.
What's new?
If this is half as popular as the previous versions - then I will be very pleased.
Happy Holiday's Everyone!
Box Shots Icons Products
Box Shots Icons Products

2010-12-14

Unsupported Hardware - Would You?

Today I came across two different products that could provide a large amount of benefit to any organization. But there is an issue with both of them.

The first one is Nexenta

NexentaStor is a fully featured NAS/SAN software platform with capabilities that meet and even exceed the capabilities of legacy storage systems. NexentaStor’s rapid development is due in great measure to the feedback and input of well over 10,000 NexentaStor users and the ongoing progress of the underlying OpenSolaris and Nexenta.org communities, each of which are estimated to have hundreds of thousands of members.

  • File system: NexentaStor can be installed as a filesystem on a wide variety of servers and server blades.
  • Software Virtual Appliance: NexentaStor is packaged as a virtual appliance with other products such as VMware’s virtualization products.
  • Community Version: NexentaStor is also packaged into a Community version for community use and for hardware and software developers that want to embed the product or integrate with the product.

The product itself seems to be very promising - with one caveat (which I will get to in a second)

The second one I came across was because of a Xsigo blog post here
(which in itself is a very informative read!)

With Xsigo virtual I/O, you can move, add, or change connectivity on demand. Configure up to 64 isolated connections per server through a single cable.

Networks and storage all connect to the I/O Director, which in turn connects to each server via a single cable (or two for redundancy). Xsigo offers both InfiniBand and standard Ethernet options for server connectivity.

Unlike any other virtual I/O or unified computing solutions on the market, only Xsigo lets you converge server I/O using the ports your servers already have.

Again this seems to be an amazing product - I have seen some demos of the product - but again with one caveat (here it comes!)

Neither of the two products are supported by VMware. They do not appear on the HCL.

Let us start with Nexenta

image

Their Support Statement and Policy

image

And last but not least

image

Looking on the VMware HCL

image

Nexenta does not even appear in the list of partners

image

Which leads me to the conclusion it is not supported by VMware.

Next on to Xsigo -

image

image

And from the the above blog post

image

The HCL does not list them either.

image

My point in the this post? Not to come down on any of the above vendors - I do not have a single bad word to say about them - not one.

But when I bet my business on a certain platform - it will have to be supported end to end - completely. And if it is not on the HCL - then I would need it in writing (in triplicate!!) that I will receive support from VMware in the event of a infrastructure issue that is because of one of these components. The fact that VMware uses the product - does not necessarily mean that I will.

I would not like to, ever, get into the situation that my production systems have an issue and I get an answer back from VMware, "It is not on the HCL - so we will try our best - but if is the 3rd parties issue then we cannot help".

Would you run your Production environment on non-HCL hardware? And if so why?
Looking forward to your comments.

2010-12-09

Veeam's Holiday Present

I would like to tell you a short story about the new present that Veeam has announced today.

image

Just under 2 months ago I sent the message above to Doug Hazelman. The text speaks for itself. This is the answer I got back from him

image 

So the Hash tags were out #*F*4*C* - which was of course #NFR4VCP

Isn't it great to have companies like this that actually think of these things before you even have the idea? and even more so - bring these things to fruition!!

The NFR licenses include 2 sockets of Veeam Backup & Replication v5 with vPower and Veeam One Solution.

This is a good strategic move on Veeam's behalf, getting the people that deal with virtualization to test drive the products, get to know them, and if the find them useful, then continue to expand their deployment

The promotion is here: http://www.veeam.com/go/free-nfr-license

Veeam

Thank you very much Veeam!!

***

(And this is completely coincidental that there have been two blog posts regarding Veeam
in two days - I swear!!)

Welcome To A New Sponsor

I would like to thank Veeam for becoming a sponsor of Technodrone

Welcome aboard!!

VMware Communities Downtime this weekend

For those of you who have not already noticed.

image

VMware will be upgrading Communities systems between 10-12 December 2010. During this time, the system will be placed in read-only mode. While we work on the upgrade, you will not be able to log in, and only public content will be available. All data will be preserved and become available again once the upgrade is complete.

This is a sneak preview of the demo site.

image

If you need urgent support during the upcoming weekend, please submit a Support Request.

2010-12-02

My Android Phone as a USB stick

A few days ago, I was presented with a situation that I had a failed ESX Server which needed to be re-installed. How and what happened is another tory for another time though.

I wanted to get some information off of the ESX host, namely the configuration and the whole /var/log partition for the purposes of opening an SR with VMware to analyze the issue.

In this particular case I could only get into single mode on the ESX host, so I did not have any network connectivity to the outside.

So I thought that I might as well use a USB stick. Well life is not so simple. The only one I had on me was a 16GB Sandisk Cruzer. Now this USB stick was formatted in Windows, and of course because of the partition size - it was NTFS.

Now the process of mounting a USB stick on an ESX host is really very simple.

  1. Plug it in
  2. dmesg to get the identity of the drive that was just inserted (for example /dev/sda)
  3. mkdir /tmp/usb
  4. mount /dev/sda /tmp/usb

It is that simple.

I needed to copy off the files before I wiped and re-installed the host.

I thought - where else do I have an available USB stick? And Bling! Light bulb!! My phone!!!

  1. Plugged it into the host.
  2. Found the identity from dmesg
  3. Created the directory, mounted the device
  4. Copied off my files
  5. umount /tmp/usb

I now had the necessary files on my phone - which I could thereafter forward to the relevant people.

Handy to have a phone wherever you go, don't you think?

2010-11-29

Welcome to VMware Alive VM

Heh? Alive VM? I have never hear of it? Well up until today neither had I.

Looking at my previous post Upgrade to Enterprise Plus promotion I clicked on the link (which was originally pointing to the above mentioned promotion) and was presented with this page.

Promotion

A quick search on VMware's site brought only these two hits

VMware

A Google search brought me to this

Google

Aha! Integrien!!

Integrien

Now Integrien was acquired at the end of August 2010 by VMware.

Now judging from what is on the VMware landing page

Get a comprehensive view of overall performance for the entire vSphere environment; health, performance and capacity. Heat maps easily locate trouble areas and mapping of virtual machines to host, cluster and datacenter combined with trended analytics enable quick and effective problem solving.

Q. What is the current offer?
A. During this promotion, when you purchase an eligible VMware vSphere product offering you will qualify to receive a perpetual license of VMware Alive VM allowing you to manage up to 50 virtual machines and one (1) year of Basic Support and Subscription (SnS) at no additional charge.

Q. What is the timeframe for this promotion?
A. When you purchase eligible products between November 23, 2010 and March 1, 2011, you will be qualified to receive a promotional code for our VMware online store which can be used to redeem the applicable VMware Alive VM perpetual license. Redemptions must be made by April 30, 2011 at midnight PST.

Q. What is the benefit of VMware Alive VM
A. VMware Alive VM is a uniquely designed product that enables the VM administrator to have a comprehensive view of the overall performance of their complete vSphere environment.

  • Indicators of health, workload and capacity
  • Heat maps to easily locate trouble areas
  • Mapping of virtual machines to host to cluster and datacenter to get to root cause analysis
  • Trending and analytics for quick and effective problem solving

    An Alive VM promotional redemption code will be emailed to eligible customers within 30 days of Alive VM General Availability or within 30 days after the eligible order is placed, whichever is later.

  • So officially the product has not yet been released. Notice the slight name change from the original Integrien brand. But the promotion is valid.

    Here are some features from the Integrien site

    AliveVM Features & Benefits

    Advanced Real-time Analytics

    • Sophisticated Dynamic Thresholding algorithms to learn the normal behavior of all time-series performance metrics captured by vCenter for all virtual elements.
    • Real-time calculation of Health for each virtual element, driven by an algorithm which assesses the underlying metric normal/abnormal behavior.
    • Real-time calculation of a Workload level for each virtual element, driven by an algorithm which assesses the resource-limited performance metrics.
    • Real-time calculation of a Capacity index for each virtual element, driven by an algorithm which projects the element’s future Workload demands

    Compelling Visualizations

    • Visualize the performance of the entire virtual infrastructure from all vCenters, not just one, complete with element relationships that tie ESXs to VMs, VMs to LUNs, etc.
    • Dashboard focus can be switched from Health to Workload to Capacity analysis; or switch from an overview of the entire environment to detailed performance information about any element with one or two mouse clicks.
    • Within minutes of installation, AliveVM’s analytics start determining the range of normal behavior for every element, so you can focus on truly anomalous behaviors that requires attention before they develop into problems that affect end users.
    • An element detail view graphically displays the Health, Workload and Capacity not only of the element themselves, but also of parent and siblings. Related storage and network devices are also shown so you understand the behavior of the element in context
    • Automatically highlights an element’s most actionable metrics, those with the most abnormal behavior in the recent past
    • View ‘mash-ups’ of health performance with related change events. This makes the impact that each change in the environment had on the element’s health obvious. For example, if a configuration change degraded the health of a particular VM, you would quickly see that on the chart, and know which change triggered the performance problem.

    Architecture

    • Available via download as either a quick and easy server installation package or a virtual appliance; as a web application there is no client installation necessary.
    • Collects all available time-series performance data, topology relationships, and vCenter configuration change events directly from vCenter. Requires only knowledge of the vCenter URL and a read-only user credential for each target vCenter to configure.
    • Install and integrate with your vCenters in minutes - no external RDBMS required, no manual configuration and no required user input.
    • Inherits the flexibility, reliability and scalability of the Integrien Alive enterprise platform for performance management.

    2010-11-22

    I Got my Ugly Mug on the VMware site

    Pablo Roesch approached me a while back to ask if he could add me to the Hero Spotlight on the
    VMware Developer Community.

    I was contacted today by a colleague in Israel who said he saw my face up on the VMware site

    The Heroes in the Spotlight are rotated, so F5 to go through them all.

    1. I am honored
    2. I am not worthy - there are so many others that should be up there.
    3. Thank you Pablo for making this happen.

    image

    2010-11-18

    Upgrade to Enterprise Plus promotion - ends soon

    Act Now to Get More Value from vSphere

    That is the title of the promotion valid for less than one month from today.

    START PROMOTION


    For a limited time, you can get VMware vCenter CapacityIQ and/or VMware View for free, depending on which version of VMware vSphere you upgrade to or purchase, making the ROI better than ever.

    If buy vSphere Standard, Advanced, or Enterprise or Enterprise Plus, upgrade to any of these editions, or purchase any vSphere acceleration kit that contains these editions
    between August 9, 2010 and December 15, 2010
    and get 15 VMs of VMware vCenter CapacityIQ, a robust virtualization capacity planning solution that helps you utilize your datacenter and desktop capacity in the most efficient and cost-effective manner. Review the list of qualifying SKUs or read the FAQs for more details.

    Buy or upgrade to vSphere Enterprise Plus
    between August 9, 2010 and December 15, 2010
    when you purchase or upgrade to vSphere Enterprise Plus and get 50 free licenses of View Premier Add-On plus 15 VMs of vCenter Capacity IQ.
    Review the list of qualifying SKUs or read the FAQs for more details.

    image

    As an additional, ongoing benefit, qualifying purchases will also be entitled to the SUSE Linux Enterprise Server for VMware operating system with subscription to patches and updates at no incremental cost. To learn more, visit the SUSE Linux Enterprise Server for VMware product page and read the terms and conditions and list of eligible SKUs.


    END PROMOTION

    Add-on packages mean you need to have a current license of VMware View.

    Regular price for 10 View Premier 4.5 add-on licenses goes for $1,500 (+$375 SNS for one year).

    Approximate price for one VM license of CapacityIQ is $75/vm (+$19 SNS for one year)

    In total - $10,785 value for the promotion.This looks like a great deal - seriously, so if you have the opportunity I would advise you use it.

    And if this is done in 3 eligible orders it goes up to $32,355. Kachinggggggg………..

    So just to clarify here - from the FAQ.


    Q. What is included in the View Premier 4.5 add-ons that will be included for free as part of this promotion?
    A. The View Premier Add-on seats will include:

    • VMware View Manager with VMware View Composer
    • VMware vShield Endpoint
    • VMware ThinApp (client and packager)
    • VMware Workstation

    Q. View Premier 4.5 add-on is limited to orders of vSphere Enterprise Plus, upgrades to vSphere Enterprise Plus, and upgrades to Enterprise Plus Acceleration Kits. Is there a detailed SKU list posted?
    A. Yes. Please refer to complete list of SKUs. <-There is a long list so please check

    Q. If I buy through a channel partner, will my orders be eligible?
    A. Orders placed through VMware resellers are eligible for this promotion. This promotion is not available for OEM partners ordering directly from VMware, except orders via Enterprise License Agreements or purchase orders where customer’s information is specified.

    Q. This promotion has a maximum of 3 eligible orders per customer. Can you define customer in this case?
    A. For the purposes of this promotion, “customer” is defined as an entity that has a unique customer name. For example, if there are 4 orders from a customer with the same name (regardless of the address), only 3 of the 4 orders will receive the View and/or CapacityIQ bundles.

    Redemption:

    Q. How will the license keys be provided?
    A. You will receive a promotion code on or before December 15th, 2010, which is redeemable for the license key at the vmware.com online store. Promotion codes must be redeemed by February 15, 2011.

    Q. Who will be the recipient of the email with the promo code?
    A. The promotional email will generally be sent to the person listed as the company contact listed on your VMware order.

    Q. What do I do if I lose my promo code?
    A. Please send a customer service request and include your VMware Order ID. We will resend the promo code that was issued to you.

    Q. What happens if I don’t redeem my code before it expires on February 15, 2011?
    A. The promo codes are only good for a limited time. If you do not redeem the code prior to its expiration, we will not be able to issue you a new code.

    Q. Can I get temporary keys for this promotion offering today?
    A. No. CapacityIQ 1.5 licenses will be fulfilled within 30 days of general availability or within 30 days of when you place the order, whichever comes later. The VMware View Premier 4.5 bundle will be fulfilled within 30 days of general availability or within 30 days of placing the order, whichever comes later.

    Support:

    Q. What Support and Subscription (SnS) is included as part of the CapacityIQ 1.5 and View Premier 4.5 add-On Bundles?
    A. When you redeem the promotion code for the View and/or CapacityIQ bundles, you will automatically receive one year of VMware Production Support at no charge.

    Q. When will the Support and Subscription (SnS) start on this View or CapacityIQ bundle? What happens after the first year?
    A. Your SnS agreement will begin on the day you redeem the promo code and will end one year from that date. Approximately 90 days from the expiration of the agreement, you will receive a quote to renew at then-current list prices.

    Q. Once I redeem the promo code and install the license keys, how do I receive support?
    A. When you redeem the promo code, a one-year support agreement will be created. The Primary License Administrator can add additional Support Administrators to the agreement via the Contract Management portal.

    Q. Who can I contact if I have questions?
    A. Please contact your VMware authorized reselling partner or VMware account representative if you have any questions.

    2010-11-16

    New Patch released - VMSA-2010-0016

    This was just released. KB 1027027 (ESX) and KB 1027753 (ESXi)

    I wanted to point out that is should deal with several issues that include a patch to solve this Host crash because of ESX Active Directory Integration.

    Here is the info from the KB 1027021

    When an user who is a member of more than 32 groups attempts to log into an ESXi host by using KVM, any one of the following issues might occur:

    • ESX/i host restarts
    • ESX/i host becomes unresponsive

    Note: With this patch, a user who is a member of more than 128 groups can access the console, but loses any group information beyond the 128th group.

    And also the 8 character password issue that William Lam posted about here

    If passwords of more than eight characters are set for ESXi 4.1 system users, the set password is truncated to eight characters, and the system evaluates only the first 8 characters of the password submitted for authentication.
    Note: To implement this fix, reset the password of the existing ESXi 4.1 system users after applying the patch.

    As always - Test before deployment

    2010-11-03

    Incorrect NFS Datastore mount

    I was contacted the other day with an issue where someone was trying to create a cluster in a box, and they were running into an issue. When trying to power on the VM with the shared, eager-zeroed-thick disk that was created they were running into an error message, saying that they did not have permission to access the disk.

    error

    I checked to see if the VM settings were correct and that the SCSI adapter was set to a different controller with the correct disk sharing options. It was.

    I enquired how was the eager zeroed disk created, and that was also done correctly. The admin logged into the ESX host and created the disk with the following command on the datastore.

    sudo vmkfstools -c 60G -d eagerzeroedthick /vmfs/volumes/IT_DBM_QT/shared/shared.vmdk

    I logged into the host and noticed something weird. The disk was created in a folder called shared but the permissions on the folder and the files were not correct. they were set with owner:group which was not root:root the way it should have been.

    From the ESX host

    drwxrwxrwx 1 root root 4096 Nov 1 14:43 .
    drwxr-xr-x 1 root root 512 Nov 1 15:23 ..
    drwxr-xr-x 1 65534 65534 4096 Sep 20 14:31 md1
    drwxr-xr-x 1 65534 65534 4096 Oct 27 11:23 md2
    drwxr-xr-x 1 65534 65534 4096 Sep 19 16:21 mich_bi
    drwxr-xr-x 1 65534 65534 4096 Oct 14 18:04 oem11
    drwxr-xr-x 1 65534 65534 4096 Sep 16 10:20 orabi11
    drwxr-xr-x 1 65534 65534 4096 Oct 3 12:09 rh5.5-m2.kickstart
    drwxr-xr-x 1 65534 65534 4096 Sep 12 12:36 rh5-dg1
    drwxr-xr-x 1 65534 65534 4096 Oct 14 16:53 rh5-dg1_1
    drwxr-xr-x 1 65534 65534 4096 Sep 12 15:46 rh5-dg1_2
    drwxr-xr-x 1 65534 65534 4096 Sep 13 11:05 rh5-dg2
    drwxr-xr-x 1 65534 65534 4096 Sep 12 15:08 rh5-dg2_1
    drwxr-xr-x 1 65534 65534 4096 Nov 1 15:07 rh5-rac1
    drwxr-xr-x 1 65534 65534 4096 Nov 1 15:09 rh5-rac2
    drwxrwxr-x 1 admin admin 4096 Nov 1 14:43 shared
    drwxrwxrwx 1 root root 4096 Nov 1 12:00 .snapshot

    Now the listing above shows that the owner was 65534:65534 which was not right.

    Now who is user 65534? I tried to cat /etc/passwd | grep 65534 but did not find anything there.

    I have the NFS datastore mounted outside of the ESX host, for administrative purposes, so I decided to check from there. And low and behold this is what I saw.

    drwxrwxrwx 17 root   root    4096 2010-11-01 14:43 .
    drwxr-xr-x 20 root   root    4096 2010-11-01 15:21 ..
    drwxr-xr-x  2 nobody nogroup 4096 2010-09-20 14:31 md1
    drwxr-xr-x  2 nobody nogroup 4096 2010-10-27 11:23 md2
    drwxr-xr-x  2 nobody nogroup 4096 2010-09-19 16:21 mich_bi
    drwxr-xr-x  2 nobody nogroup 4096 2010-10-14 18:04 oem11
    drwxr-xr-x  2 nobody nogroup 4096 2010-09-16 10:20 orabi11
    drwxr-xr-x  2 nobody nogroup 4096 2010-10-03 12:09 rh5.5-m2.kickstart
    drwxr-xr-x  2 nobody nogroup 4096 2010-09-12 12:36 rh5-dg1
    drwxr-xr-x  2 nobody nogroup 4096 2010-10-14 16:53 rh5-dg1_1
    drwxr-xr-x  2 nobody nogroup 4096 2010-09-12 15:46 rh5-dg1_2
    drwxr-xr-x  2 nobody nogroup 4096 2010-09-13 11:05 rh5-dg2
    drwxr-xr-x  2 nobody nogroup 4096 2010-09-12 15:08 rh5-dg2_1
    drwxr-xr-x  2 nobody nogroup 4096 2010-11-01 15:07 rh5-rac1
    drwxr-xr-x  2 nobody nogroup 4096 2010-11-01 15:09 rh5-rac2
    drwxrwxr-x  2 admin  admin   4096 2010-11-01 14:43 shared
    drwxrwxrwx 10 root   root    4096 2010-11-01 12:00 .snapshot

    The owner was nobody:nogroup.

    I then asked the storage admin to please check the export on the NetApp filer was defined and got this back

    /vol/DBM/DBM_QT        -sec=sys,rw

    I remembered from a while back in my first VCP course that ESX hosts have to have root access to the NFS mount in order to work.

    Changed the export to:

    /vol/DBM/DBM_QT        -sec=sys,rw,root=1.1.2.0/24:1.1.3.0/24

    I now tested the creation of a new VM from the vSphere client and verified that the permissions were now created with root:root.

    Strangely enough the ESX host was able to mount the datastore, machines were working as well.

    I did however still have to change the permissions on the folders and files that were already created on this datastore while the mount was not correct. This I did in order to prevent further problems and rectify the incorrect permissions

    chown -R root:root /vmfs/volumes/IT_DBM_QT/*

    Of course this was all done with the VM's powered off!

    2010-11-02

    Get-vSphereClient Function

    As was posted on the new KB article 1029925 the vSphere Client is no longer packaged with current builds of ESX and ESXi. However it is packaged with vCenter, so you can still get it from the web interface of your vCenter server.

    Since they have posted a direct URL to the download I decided to wrap this in a PowerShell script
    Get-vSphereClient.

    Function Get-vSphereClient {
    <#
    .SYNOPSIS
     Downloads the vsphere 4.1 Build 258902 Client
    .DESCRIPTION
     As was posted on the new KB article 1029925 the vSphere Client is no longer 
     packaged with builds of ESX and ESXi. This function will download the vSphere 
     client from VMware's site
    .NOTES
     Author: Maish Saidel-Keesing
    .PARAMETER Url
     The url for the .exe file
    .PARAMETER Path
     The location the file will be saved
    .EXAMPLE
     PS> Get-vSphereClient -url http://vsphereclient.vmware.com/vsphereclient/2/5/8/9/0/2/VMware-viclient-all-4.1.0-258902.exe -path c:\temp\viclient.exe 
    #>
    
    param (
    $url="http://vsphereclient.vmware.com/vsphereclient/2/5/8/9/0/2/VMware-viclient-all-4.1.0-258902.exe", 
    $path="c:\temp\VMware-viclient-all-4.1.0-258902.exe" 
    )
    
    ##Create Web Client 
    $client = new-object System.Net.WebClient
    $client.DownloadFile($url,$path)
    }

    Hope you can make use of this.

    Jase McCarty posted an article on on how to change this behavior on his blog a few months back as well.

    List of Free Ebooks

    I came across this list of free eBooks that Microsoft has released as of late.

    Windows 7 troubleshooting tips

    Deploying Windows 7, Essential Guidance

    image

    Introducing Windows Server 2008 R2

    image

    First Look Microsoft Office 2010

    image

    Perhaps you can find something useful here.

    2010-10-31

    Can We move to only one Physical CPU?

    I started reading Eric Siebert's book this afternoon - Maximum vSphere: Tips, How-Tos, and Best Practices for Working with VMware vSphere 4 (which is a great book - highly recommended!!), and for some reason during the part that he was speaking about licensing and the features, a thought crossed my mind.

    I put out a feeler this evening on Twitter with this question:

    Quick twitter poll - what is your average CPU usage on your ESX hosts? <25% - <50% - <75% - >75% - Interested to hear...Sat Oct 30 18:43:27 via TweetDeck

    All the answers I received all pointed to the same conclusion.

    The constraint that almost everyone hits first is RAM, not CPU. Some admins cannot expand on the amount of RAM in their server, because the cost of the bigger DIMM's are too high, and there are not enough slots left in the server. Which leaves them with servers that are nearing memory capacity, but not not anywhere close on utilizing the CPU power of the server.

    Many people are purchasing dual-socket servers for redundancy or because of the fear of not having the server perform well enough.

    From my own environment I can say that my hosts are utilizing around 30% of their CPU, with 2 Quad Core CPU's. And from the answers I got tonight on my question above - the results are pretty much the same.

    Now perhaps a sacrilegious thought. What would happen if we only used one physical processor in a server?

    Today we are talking about a six or eight core processors and this number is rising. The amount of cores available are more or less the same,  as what the majority of people are using today, 8 cores - 2 x Quad-core processors.

    Now you might ask, but here I lose the redundancy. This could be true, but how many of you have actually lost a CPU due to malfunction in a server? I personally have not. Ever. I would also suppose - that if a physical CPU barfs on you during a production workload - then it will not be pretty. The VM's that were running from that Processor will obviously kill over and die, but I suppose the rest of the host will not be happy either. From my experience with faulty memory, you are more likely to crash the whole host with a PSOD than having the host function with one DIMM less. I guess that with a CPU - it will probably be the same. So having redundant CPU's does not really cover it. I could be wrong here, and if so I would appreciate your feedback with more information.

    Now I am sure there are other implications here, regarding the spread of memory and load over the two channels from both processors, and I am also sure that there are other internal ESX performance implications as well. So it is not a simple matter.

    How will this change the game though? Well it will cut costs - in two ways.

    1. Licensing. ESX licenses are now counted per processor, and not per sets of 2. Removing one processor, will lower ESX host licensing costs by half.
    2. Server hardware. With one processor less, you are able to cut costs on each server.

    So are we destined to run only a 1 Socket ESX host? I would interested in hearing your thoughts and insights on this one.

    2010-10-23

    Journey to VCDX - Not complete - YET!

    10 days ago I completed my VCDX defense in front of a Panel of some the top professionals and technical people in VMware. This  has been the completion of a journey, a path and quest to achieve the VCDX certification.

    But let us roll back a bit - approximately a year back.

    VMware announced that there would be a new certification path that would be a level above the current VCP certification.

    There was a great amount of buzz around this when it launched and people were all gung-ho to start and get this new cert.

    Until.. They saw what was involved. And then a large number of people dropped it.

    VMware Certified Design Expert - that is the abbreviation. 4 letters but oh it is so much more.

    So let's start why I decided to pursue this certification.

    Let's face it. There are approximately 56,000 VCP certification holders. That number becomes less for those who hold both VCP3 and VCP4, and even less for those who hold VCP2, VCP3 and VCP4. I am not a certification chaser, not by a long shot, usually I only update my Certs if I have to or find additional benefit, it was the same with MCSE 2000, and MCSE 2003. I have not pursued the 2008 certifications, because I did not need to, but more so - I do not see the benefit of doing so at this time. Being one of 56,000 individuals in a population of almost 7 billion is an amazing achievement.

    But I aspired for something more. Doing the regular day-to-day admin work was not enough. So you start to look for more interests, you go into new subjects, you continuously learn new things.

    Now this certification was something that caught my eye. This was not you average Admin work. The most interesting thing that I saw, was that you could not study from a book for the exams. the Exam blueprint was based on experience and in-depth knowledge of VI3 products and the surrounding infrastructure. There were no braindumps, no cheat sheets. This is either something that you know or you do not, based primarily on you own experience and knowledge.

    I made a decision, that I would try to pass the Enterprise Admin Exam, and see further. I passed

    And then I decided that I would try and pass the Design exam thereafter. And I passed

    I was now at a crossroads. The last two parts was submitting a design, and defending it in front of a panel of architects. Now I kid you not. No matter how confident you think you are (I have still some work to do on this), no matter how clever you think you are, no matter how good you think you are, this is not an experience to take lightly. I knew that I would have to travel overseas for this part, and that would entail securing the correct funds and time to complete this. I managed to secure both.

    The application is one of the hardest things I have ever done. In my life.

    When designing an infrastructure, it is more than just consolidation ratios. It is more than how much RAM each ESX host should have. It is more about than how many IOPs you will need to support you environment. It is more about than how many ESX clusters you will have, what level of DRS, what will the failover settings be. I kid you not, each and every one of these above are extremely important and should be a fundamental piece of your design. But the whole idea of being an architect (which is what VMware is looking for - IMHO) is all of the above, but also to understand how the your business works, what are needs, what are their requirements, how you plan to fill those needs/requirements. And how every decision you make can/will affect some other part of the environment.

    You need to be able to see the big picture. You need to cater to the needs of the customer. And even more so, be able to justify to mainly to yourself, and in turn to the customer, why this option is better than the other? What will happen if you choose one over the other?

    The application is brutal - seriously. It is not a clear document saying:

    • Explain in 10 words or more what you design is.
    • Present a diagram of your virtual machine layout.

    It is open to interpretation. There are a decent amount of deliverables you have to provide - but there are no instructions exactly what has to be in these deliverables. I personally find it extremely difficult to provide an answer to something without knowing the what is expected of me. But in essence here is the good thing.

    Let me give an example.

    You are asked to build a tower out of cards that will be 7 levels. The end result is that it will have to stand on its own without support for 10 minutes.

    So you might ask - what I am allowed to use? How many cards, any geometrical shape? will there be wind that will blow these cards down? Am I allowed to use anything else besides cards? All of these are legitimate questions which will help you in your planning and design of your tower. Some of the answers you receive and some of the answers you do not. Now of course there is more than one way to accomplish this. But in the end what you will have to to I make sure that you tower stands and you will have to explain why you chose what you did, and what were the implications of choosing A over B.

    So for example you might start to research what is the most steady geometrical form, create a schematic on how to build it, measure to the millimeter, what card has to go where. Then you build, document, test etc.

    On the other hand you might just start and see what works for you. stick the cards with Cello tape to one another, staple the bottom level to piece of wood and build. You then document the process and reproduce it.

    Which solution was correct you may ask? I think both, each in their own way. Both provided the end result. Both were documented, both are reproducible.

    Sometimes, giving you a detailed spec of exactly what is required - suppresses creativity. We all work with both sides of our brain, the right for creativity and the left for analytical thinking. Some use one side more than the other. There is usually more than one way to solve a problem.

    For the VCDX - I think that the journey is the most important part. How you got to decide that you will have 24 VM's per LUN? Why you use VMFS and not NFS. And if you were to change that decision to something else - what would the implications be. I have learned a lot from this journey, it has helped me grow intellectually, technically, professionally and personally. And for that I am extremely grateful.

    Today I received my answer from the panel, and unfortunately the answer was no. There were a few points that were noted that needed improvement.

    I would like to thank the Panel members for their time and effort put into each and every defense. I gather I will be seeing you all when I re-submit for a defense.

    I do not remember where I once heard this but it is so true. "If you do not have the courage to fail, then you will not have the courage to succeed."

    I see this as a setback, and as a learning experience. Perhaps I took too much on my plate. Too much going on at the same time. And not enough time to concentrate on each thing properly. Or then again, maybe not.

    When I finished the defense I felt immense relief. I felt that I had given my best in the process. I felt the interaction was good between the panel and myself.

    I think that I am the first (or one of the first) public figure(s) in the virtualization world that has not passed the defense. Nothing at all to be ashamed of. The success rate is not high, not at all.

    I have decided not to re-submit for a defense for the VCDX 3 track. The time left to submit a design is not something that I can complete (at least to my satisfaction) before November 22nd.

    I am going to give myself a rest for a few months and start out anew for the VCDX 4 track. I need to complete some other projects that I am currently busy with. Yes I know that it will mean that I need to complete the exams again, but again that I see as part of the learning experience.

    So back to the blogging, back to the technical troubleshooting, back to answering questions on the forums.

    And that was the first part of my VCDX journey..

    2010-10-21

    VMworld 2010 - or was it Disneyworld?

    A week has passed since the last day of VMworld in Copenhagen. And I owe myself (and you as well) a roundup and summary of the event

    Quite a few roundups have been published over the past week. I will not repeat what has been said in the other posts, but I would like to add my following comments.

    Which kid does not like going to Disneyworld? So I guess you can say the same thing, which virtualization enthusiast does not like like going to VMworld?

    There are very few events or places that you can go and find a load of people who are interested in the same thing you are, who understand what you are talking about, who will crack up laughing at the funny slogans on the splunk tshirts. So thanks VMworld.

    There are many opportunities we all have to play with new technologies. But we hardly ever have the privilege to have everything prepared for us, the same way we provide for our end users and customers. That is why I think that the labs were such a hit. Of course the experience, the diversity, and the performance of the labs was amazing, but for me to go and and have a a vCloud Environment all ready for me with the click of a mouse, without me having to go and find hardware, storage, licenses and time to set it up, was a treat. So thanks VMworld.

    The social networking. It was great finally meeting a great amount of people that I have chatted with, tweeted with, DM'ed and read their blog posts over the past year or two. It was an honor to meet you all. The conversations were great. The live troubleshooting sessions were great. The ad-hoc meetings were great. So thanks VMworld.

    Meeting and talking with the VMware Product Managers / Architects. Meeting and talking with the Vendors, the Directors, VP's, Architects and big-wigs. I have mentioned it perhaps before, but I am continually amazed at how accessible you all are. I sometimes think that maybe it is me who is well connected, but no it is not! It is all of you who are willing to listen, to help with a problem, to make the change when necessary. So thank you all and thanks VMworld.

    I heard and read many a blog post about what is the important part of VMworld. Is it the sessions? Solutions Exchange? Labs? Maybe the Social Networking? I think it is all of the above,

    I tried to divide my time between all of the above. I gave less of an emphasis on the sessions and Solutions Exchange. I picked and chose my sessions carefully. and only participated in 3 in total. I chose them because of the topics that I found appealing, the rest of the sessions that I want to see, I will download in the near future. I tried to get take some labs with products that I will hopefully be evaluating in the very near future to get a quick feel for the product. The few products I spent time with at the Solutions Exchange were ones that I felt would benefit from in the upcoming months, mainly with insight into the performance and analysis of my environment.

    It was a blast, I think it was one the highlights of my entire year. It makes all the time spent blogging, on Twitter, writing and on email so worthwhile.

    So thank you VMware for having VMworld and thank you all for making it the amazing experience it was.

    See you next year!

    VMworld 2010 Copenhagen - vExperts Panel

    I wanted to share with you a short interview that John Troyer held with Alex Mittell, Erik Scholten and myself during the Show last week.

    I apologize I did not manage to trim the video properly, so please scroll to the 3.00 minute mark to get to the beginning.

    Hope you enjoy.

    2010-10-18

    Getting Average Usage from your hosts - PowerCLI

    I was asked to provide a small report on what the average RAM usage was on each host over the past 30 days

    I forgot to mention the code was inspired by the session that Luc Dekens and Alan Renouf gave at VMworld.

    Of course you could do it with mouse clicks – but man I love PowerCLI. 10 lines of code. less than 3 minutes and BAM!

    $start = (Get-Date).AddDays(-30)
    $finish = Get-Date
    $metrics = "mem.usage.average"
    
    Get-Datacenter | ForEach-Object{ 
     Get-VMHost -location $_ | Sort-Object Name | ForEach-Object { 
      Write-Host $_.Name 
      $stats = "" 
      $stats = get-stat -Entity $_ -Stat $metrics -Start $start -Finish $finish 
      Write-Host "$($_) average Memory Usage over the last month is: $("{0:N2}" -f ($stats | Measure-Object -Property Value -Average).Average)%"
     } 
    }

    1-2. Set start variable with the date I need (30 days ago) and finish with today's date
    3. This is the metric I am looking for.
    5-6. Here we get each Host, I divided it up by datacenter/cluster to show that all hosts are (should be) balanced and using the same amount of resources and we perform the action on each host.
    7-8. First clear the variable, from any previous runs, and set the stats variable with the information I want.
    9. Here I used the .Net format for numbering "{0:N2}" -f otherwise I received a number which had 10 digits after the decimal point (which is pointless .. ;) )

    And I got this output

    Dev
    esx2.maishsk.local average Memory Usage over the last month is: 82.46%
    esx3.maishsk.local average Memory Usage over the last month is: 79.20%
    esx5.maishsk.local average Memory Usage over the last month is: 79.04%
    esx6.maishsk.local average Memory Usage over the last month is: 72.57%
    esx7.maishsk.local average Memory Usage over the last month is: 77.97%

    Of course you can do this with any metric as well.

    Time to add more hosts to the Cluster (me thinks)….

    2010-10-14

    VMworld - Day 4

    Today was the day of packing up. Early check-out from the hotel.

    I headed over to the Labs early to get a lab done. The labs were already busy at 08:10. The Wyse thin-client on each and every station was what connected you to the labs and from there further into the cloud to the remote datacenter

    image  image

    The lab I completed was LAB18: VMware vCloud Director - Networking. This was the 3rd in the series of Cloud related related Labs. I think I have said this before. The hardest part of understanding vCD is going to be the Network. I think the rest of the concept is really easy - especially if you have used Lab Manager before. The different kinds of networks and segmentation of these networks is much more complicated than the VIAdmin has been used till until now - dealing with the vSwitch and the dvSwitch. It seems that In addition to having to administer the Hardware, the virtualization layer, the Operating Systems, and as of late Storage is becoming a more integral part of a VIAdmins day to day job, we will now also start to design/manage the network in our cloud environment - or at least have to understand how it works, before selling it to your customers, or implementing in our own Environment. The lab had some quirks and did not run smoothly. And at one time, I had Duncan Epping and two other proctors (I apologize I did not get your names) around me trying to to solve the issue, which they did, which allowed me to complete the lab.

    I finally got Duncan to autograph both of his books(one day they will be worth millions $$$$ on EBay)

      vSphere 4.0 Quick Start Guide and

     image Foundation for Cloud Computing with VMware vSphere 4

    After the Lab I moved over to the Blogger Lounge where at one time, where there was such a concentration of talent, knowledge and good moods - It was great. Scott Lowe was kind enough to give me a Spousetivities shirt to take back for the missus. On a whole, besides one or two sessions, I think the largest benefit I gained from this conference was interacting with people. But more on that in an upcoming blog post.

    I wanted to go back into do another lab, but the lines - man the lines!! 20 minute wait just to get in!

    image

    I came back to the lounge and had a great discussion with John Arrasjid, about politics, about cloud, about religion, about home labs, about the VCDX process. It was great.

    I was on my way to a VDI performance assessment session on the other side on the Center, I bumped into Scott Herold, and I went past the session (TA8133 Best Practices to Increase Availability and Throughput for VMware) with Vaughn Stewart and Chad Sakacc. I decided to sit in this one instead. The room was packed! Completely!! I have never seen either of them present before (besides hearing recordings). Both of them are amazing presenters, really - they both know their stuff, have a great amount of knowledge. A great mix of detail, technology and entertainment.

    It was a really good session on storage performance best practices. Nothing that I have not heard or seen on their joint blog posts, but hey - you gotta go and see the "Frenemies" in action.

    After the session, I went up to say hi to both of them, and had a lengthy discussion with Vaughn about NetApp, storage, vision and what is to come. It was a great discussion.

    I am constantly amazed how accessible the top people in the industry are. I mean if I were to compare it to the CTO of my current employer or other companies for that matter, I am not sure how many of them are available for speaking to the end user. You obviously have to take and make the time to allow this to happen, and I am not only talking about the two gentlemen above, This is something I constantly noticed over the last few days, at all levels, I spoke to VP's CTO's and Principal Architects.

    And really who am I? An Admin. So my hat off you all. You are what makes this industry exciting, enjoyable and a better place for us all.

    Ok enough philosophy.

    I went to sit in one more lab. vCenter configuration Manager - This is a whopper of a product - but I must admit, the amount of clients that will actually get to implement it are not large. The product has a great amount of functionality which assures your Environment is compliant to a baseline that you have defined. It can also revert changes back to the compliant profile if a change has been made. Not only for Windows machines, but also Unix, Linux and ESX hosts as well. Which means you can define a profile for a host/vm. If someone (an ambitious power user who has the correct rights) decides to bump up the amount of RAM on the VM for better performance - but is not allowed to - then if the item goes out of the compliant profile, you can be alerted, and if you would like you could revert that change back automatically. I have note really looked at the requirements for the product regarding resources, or cost, but it should be something to keep you eye out for.

    One last walk around the Solutions exchange - one more t-shirt from Splunk. I finally got to see a real live UCS - and I must say that I am impressed, the simplicity of the product makes it all worthwhile. now just to convince upper management.

    A round of goodbyes to all the people I have met that I previously only knew by their Twitter handle and I started my journey back home.

    And that was day 4 - the last day of VMworld Europe.

    One more blog post to come to summarize my experience.
    And no - I - did - not - win - an - IPAD! Smile

    VMworld - Day 3

    Day 3 started with me being a bunch of nerves and getting up at the crack of dawn. I would like to lodge an official complaint with the Danish Hotel Association (If that even exists)

    You should not have two hotels named the same in the same city

    First Hotel Vesterbro // COPENHAGEN and the First Hotel Skt. Petri // COPENHAGEN. Not being Danish - can you see how this can be confusing - and they are located within 1.5 Kilometers of each other.

    This caused confusion this morning which almost made me late for my VCDX defense. I will talk about that in a separate post.

    After a short walk through town, and back at the Bella Center I felt so relieved after the defense was over - so it was back to the conference.

    A session with the support services of VMware that I was invited to (but left before it was finished) - after receiving some interesting information which I shared on Twitter

    8 character password issue is 4/12 of top support issue in vSphere 4.1 - http://bit.ly/bSn8t2Wed Oct 13 12:47:34 via web


    http://bit.ly/d1zQCc 5/12 top support call Host Crash because of ESX Active Directory AuthenticationWed Oct 13 12:49:10 via TweetDeck

    Solutions Exchange was still buzzing - it was quite amazing how much you do not notice the noise when you are down on the floor.

    imageimage

    I then joined ALT2004 Building the VMworld Lab Cloud Infrastructure - Dan Anderson. Absolutely amazing!!

    I so enjoy these sessions which are given by extremely technical people, who do the every day job that I do as well, and it is great to hear how they did it, what were the issues they came across along the way and how they dealt with the issues.

    A quick summary.

    • A large part of the lab infrastructure is running on Nested ESXi
    • This was the Infrastructure in San Francisco

    image

    • And this is what was set up in Copenhagen

    image

    • There was another full datacenter set up in the Bella Center itself - but was not used. (redundancy)
    • Storage used was iSCSI, NFS, and FC
    • Cloud Lab is an internal product by two individuals at VMware employees who developed it as Dan put it "on their second day job"
    • They were running some 64 bit VMs under the Nested ESXi hosts. Something which is not possible to day. (and of course he did not tell us how it was done)
    • Dan's Manager "volunteered" him for the job of creating a Whitepaper on the whole process.

    After the session the Q&A continued further in to what exactly that limitation is that does not allow a 64bit VM to run under a nested ESX.

    If you do download the sessions, this one is a must.

    After this session, I returned back to the blogger lounge - to be asked to join a short 1 minute piece on the VMworldTV channel with John Troyer.- I will post the the link to the video when it becomes available.

    I sent some time at the VMware party, music, drinks and food were a`flowing.

    And that was Day 3.

    And I still do not have an IPAD. (but Joep Piscaer does!)

    just won an Apple iPad 16GB WiFi at the #VCE booth. Thanks #EMCWed Oct 13 08:45:08 via TweetDeck

    VMworld - Day 2

    My day started at 08.00 with two labs that. The labs were nice and quiet.

    LAB17 Lab: VMware vCenter™ Orchestrator

    This I think is one of the most hidden secrets and most unused featured that is bundled with vCenter. In essence this could enable you to basically do almost anything you want in your Virtual Infrastructure. But the issue is that the documentation for the product was, IMHO, severely lacking until the current release and besides a few how-to's on how to set it up - the knowledge is practically non-existent.

    I would like to point you all to the VCO Team site who have been releasing amazing content on how to make better use of the product.

    The first part was the installation process of Orchestrator which has been simplified even more than previous versions.

    The Lab continued on how to create workflows to deploy VM's and move VM's from one location the other.

    I would have liked to see a more advanced level for giving more insight on how to configure and create workflows and how to present them to the end user through the web interface, maybe at VMworld 2011

    LAB20 Lab: VMware vCenter™ vShield 2.0

    I was extremely impressed by the product. The lab gave a good overview, precise explanations and the enough amount of hands-on to see an example on how to load balance or NAT a VM, amongst other things.

    I started to walk around the Solutions Exchange floor. SWAG, SWAG SWAG!! Mike has started a worthy cause of the The SwagBag Competition.

    I had two very interesting demo's. One with the VP of Xangati who have what seems to be a very promising product. The second was a demo of Splunk. They actually had the funniest t-shirts. Seriously! This is something I would like to investigate further after the show.

    I then spent some time in the Blogger lounge where it was great to put a real face to the Twitter handles that we all use. Great discussions about all sorts of things are happening the whole time. @jtroyer (so much taller than I expected) is busy with the VMworld LIVE video feed and interviews throughout the day.

    A good part of the day was spent in the blogger lounge @h0bbel was kind enough to be the blogger lounge photographer.

    image

    I then went to the BC7803 Planning and Designing an HA Cluster that Maximizes Virtual Machine Uptime - which was packed solid - Duncan Epping 

    imageimage

    Back to the blogger lounge there after

    image

    Then we were invited to a the CTO event with Dr. Stephen Herrod. There were some great conversations that I had with Scott Lowe, Wil van Antwerpen, Bob Plankers - just to name a few.

    I also got a chance to have a short chat with Dr. Stephen Herrod, to express my appreciation for hosting the event and giving us the opportunity.

    I gave a pass on the several parties and dinners that evening and spent the rest of the evening in my hotel, finalizing my VCDX defense presentation.

    And that was Day 2!

    And No - I still did not win an IPAD