Platform Products

Virtualization Platform Products

Silence!!! Serious Small ESXi hardware, awesome for you home lab :-)

While I am playing on making cheap, cool hardware gadgets for VMware, others are focused on finding the best possible hardware to run your ESXi (or other hypervisor) on. Alex from the UK has made some real awesome progress on this in the past few days. Using a newly released Intel NUC, which is a new barebone system for 260 EUROS (inc VAT) ex memory, but capable of max 16GB he found a way to run ESXi v5 on it. Not only is this a good enough system for home lab; 16gb RAM, 2x Intel I3 Cores, 1x 1GB LAN, and the option to add extra via mini-PCI-express cards (~43 euros) it is also Quiet!!! (like in silence), which probably for most home labs will be awesome. My servers have to be in my garage (out side of the house) due to their noise. O and from the energy conscious geeks, it also just uses a few Watts 🙂

The total system with 1nic and 16GB ram comes down to 348 euro! (inc VAT). This put having 2 physical lab nodes into the possibilities of many virtual geeks out there… and without pissing of their girl friends /wifes (at least noise wise).

Read here the full article by Alex how he got everything up and running.


Network Acquisitions are hot :-)

Well well, what is going on in the virtualization landscape. I guess more compute virtualization challenges are solved, so it is up to the next challange, network virtualization. Today IP addresses really pose a big problem in creating an agile environment, as it represents both identity and location.

To help start solving some of these network challenges, VMware and Oracle both announced big acquisitions in the networking space.

VMware will acquire Nicira, who is specialized in software-defined networking. Oracle will acquire Xsigo, who as well is specialized in software-defined networking 🙂

More about both deals:
VMware press release
Oracle press release

Is it a coincidence that both virtualization players are getting into software defined networking? Of course not, as I said, most problems around compute virtualization are solved today, while networking is still very much old-school. If we want environments where virtual machines can move from on-premise to off-premise, from one datacenter to another, if we do not want a 1000 cables sticking out of our highly consolidated servers, we need to get the networking on the virtualization bandwagon 🙂

When both companies will integrate these solutions into their stack it will truely be a win for all their customers, who will gain a lot of networking flexibility because of it.

Can Microsoft make themselves even more ridiculous?

I was browsing today the internet and came across this ‘dumb’ video. I guess Microsoft is trying to fight the competition, so they made this ’10 myths about VMware” video. Silly of them to not use any accurate information or counter arguments.

Microsoft Mythbusters: Top 10 VMware Myths

I feel very tempted to shoot down all their 10 points, but should I bother? Point 1 and 2, claim that they have live migration and clustered file system, but Microsoft’s definition of “having” can I guess include future stuff. I am of course an alien, but in my english classes i learned there is a difference between “have” and “will have”.

Being a techie my self, one point I do want to address, which is memory commitment, why does Microsoft just want to refuse to see how it works. It does NOT work by not allocating unused memory! It scans memory pages to find duplicate pages, so if you run 10x basic windows OS, and yes that memory can be actively used, it will only store it once in memory. So memory commit feature ‘compresses’ the memory, and if you do not do it in extreme, you can do it without any serious punishment, while still saving memory.

OK, one final one, and this has nothing to do with technical stuff. Mister Microsoft claims that it is silly to compare based on cost per application / virtual machine, Microsoft and VMware prices are based per machine, so that is how you should compare????? Hello Microsoft, get real!!! It is like saying, when you buy a car, only look at the price of the car, do NOT look at the miles per gallon (KM per Litre). If Microsoft keeps up that strategy, they will end up like all american car manufactures, not selling a single car anymore, while the efficient Japanese car manufactures laugh their asses off.

Competition should be great, it keep manufactures sharp and on the edge, but for some reason Microsoft just plays silly. I hope they will actually make some good features at some point in time, so that the market moves forward instead of playing ignorant.

It is good to see that this nonsense does not get unnoticed, have a read of this blog post, it exposes how Microsoft IT looks at their own virtualization product.

Cisco unveils their new Unified Computing System

After many weeks of speculation, Cisco today announces their new Unified Computing System.

The system will come with VMware’s virtualization software as VMware also today confirms that they have made an OEM agreement with Cisco.

Get full control over your ESXi Server

As you probably know ESX 3i is the embedded version of ESX, mainly run from usb flash disks build-in servers from HP, Dell, IBM and FSC, but you can also install this on your own servers. I really love this version as it is so easy and simple and adds extra (and less) functionality to ESX. The main thing about 3i is that the Red Hat service console has been removed and replaced by a small busybox linux version. To keep this service console tiny, 3i comes with build-in hardware monitoring, something a normal ESX server does not have (wish it did). When connecting to a 3i server with your VC client you can see fan speeds and temperatures in your system and you can of course put alarms on them, in case something breaks.

By default the 3i server gives us no access to this small service console, but there are some down sides to this, as certain things can only be done from the service console 🙁 and until VMware builds those missing functionalities into the VC interface you still need to get yourself console access. In the beta versions of 3i there was an option in the advanced settings, but this has been removed in the final releases. VMware has published a knowledge base article that describes how to get console access (this is not officially supported by VMware). See the KB Article. In short you just hit Alt-F1 and then type in ‘unsupported’. The article also explain how to disable this Tech Support Mode, in case you want to.

Getting console access

Of course you do not want to be in the server room all the time and your server might not have some kind of remote access facility, so the first thing you probably want to do is enable SSH access to your server. (Thanks to Lee for this information, and again this is not supported by VMware)

  • edit /etc/inetd.conf (using vi)
  • remove the # (remark) sign in front of the SSH line
  • kill and restart the inetd process (or just reboot your server)
  • So why do you need console access?
    There are multiple reasons why you want to have console access. One of the most common reasons is that you might want to change HBA driver options. To get better performance out of your Qlogic or Emulex HBA most people increase their queue depth (see for instance VMware 100.00 io blog post). To find out what options you can set on your drivers you can use the command vmkload_mod:

  • display your loaded drivers: vmkload_mod --list
  • display parameters for your driver: vmkload_mod -s /mod/your_driver
  • Another reason why you definitely still need console access (unfortunately) is if you want to use thin provisioned virtual disks, a new feature in ESX, but for some reason not exposed in the VC interface yet (i think because it is still experimental supported). Any virtual disks created via the normal interface are pre-allocated disks, so a 100GB virtual disk will use 100GB on your VMFS. With thin provisioned disks the actual virtual disk file will start very small and only grow when you will actually need the disk space. Note: this only works for Virtual Disks on iSCSI and FC SAN, not NFS!

    With the vmkfstools command you can create new virtual disks as ‘thin’ disks. After you created the disks, you can then use the normal VC interface to add that disk to your VM.

    Creating a 100GB thin disk:
    vmkfstools -c 100G -d thin /vmfs/volumes/san_vmfs/my_vm/thin_disk.vmdk

    If you already have a pre-allocated disk, you can convert it to a thin disk as well. Well it is not really converting, but creating a new copy as thin disk. After you have done that, you will need to remove the old disk from the VM and add the new converted copy.

    ‘convert’ copy a existing virtual disk to thin format:
    vmkfstools -i /vmfs/volumes/san_vmfs/my_vm/large_disk.vmdk -d thin /vmfs/volumes/san_vmfs/my_vm/new_thin_disk.vmdk

    Well I am sure there are more reasons why you still want to ‘play’ on the service console, as I am born with a commodore 64, I always want a command prompt 🙂

    Is Microsoft really that stupid…

    Or are they just playing for stupid???

    Today I was reading a blog post on the Windows Virtualization Team Blog, on their 3rd story about Microsoft’s Quick Migration and VMware’s VMotion. It is just too funny to see how Microsoft is dealing with a lack of functionality. According to their research customers are not changing their patterns, even when they have the option to manage physicals hosts in their environments during the day without any form of downtime for their Virtual Machines, they (according to Microsoft research) still want to delay physical server maintenance till the evening hours. Thanks to this research Microsoft is claiming that being able to move virtual machines around physical servers is really not something useful to have. The thing that really amazes me is that I see Microsoft (and others) often make assumptions based on the old-fashion one-app-per-one-physical-box time era. Please wake up! Virtualization is a Disruptive Technology… You shouldn’t be working in the wonderful world of IT is you are not willing to move forward (at least in my opinion). I do know for a fact lot’s of customers are using VMotion (or live migration) in production… sometimes helping them to do server maintenance… and really daring and probably red bull drinking extreme sport IT managers even try this during the middle of the day…

    But guess what… the ability to move virtual machines while running without downtime solves MORE then just silly hardware maintenance!!! It creates an environment where workloads can actively be rebalanced so you can better serve your application needs and the end-users using them. And I am sure this new innovative (well at least 3 years ago when it came out) technology is being welcomed by many extreme sport red bull drinking IT managers  It is even better to see that technology is not sitting still again… Storage VMotion, a recent released technology, now-a-days also allows you to dynamically balance storage across different tiers of storage as well (No, Microsoft, this is not just a feature to help prevent storage hardware maintenance). With a bit of patient we will get our hands on the continuous availability technology VMware has demonstrated at VMworld events where a VM can be continuously replicated across a simple network cable, so that when a server does die on you, the VM will keep on running without a hickup… Long live innovation… for the IT people preferring to maintain their hardware in the evenings… I guess you must not live in a neighborhood like mine with a nice strip club 

    P.S. Just to clarify, before everyone flames me, yes I work for VMware… But I work for technology, not for a company, so I choose to work for VMware because I like their technology (I am free to switch jobs any time I like). No one in VMware asked me to write this and feel free to comment if you disagree or agree  you might even convince me to apply for a job at Microsoft 

    The Microsoft Article

    VMware Announces VI 3.5

    Today VMware has announced VMware Infrastucture 3.5 🙂

    Key new features:

  • ESX Server 3i
  • Guided Consolidation
  • VMware Update Manager
  • VMware Distributed Power Management (Experimental)
  • Read more here about the new features
    Read the press release

    Hacking your ESX 3i Server

    Well all my ESX servers at home are now migrated to ESX 3i and I have to say what a great concept. Just stick in a memory stick and you are up and running. After playing with it for some days now I wanted to share some tricks with you.

    First ESX 3i is based on the new ESX platform and therefor now supporting SATA drives 🙂 Yahooooo! I can run ESX now properly and native on my laptops. I have tested this with my IBM X60 (Centrino Dual Core) and Dell Latitude D410 (Pentium M single core) and on both it works fine. It sees my drives and I can format a VMFS partition on it and it supports the build-in NICs in the laptops.

    Getting a command line
    The cool thing about ESX 3i is of course that it is small and no longer has the Service Console.. but this means also NO command line on the ESX server anymore 🙁 But being a techie, I love command lines! And after some searching I did find out you can still get a command line on the ESX 3i Server. Sure it does not run a full blow Linux, but most of the VMware commands like vmkfstools and esxcfg-* are still available.

    So how do you get this command line? Well easy, follow these steps:

  • Power on your ESX Server
  • On the Boot menu, hit shift-O, you will get an advanced option line now at the bottom
  • type in the advanced option: shell
  • Continue now the boot of the ESX server, after it is powered on hit Alt-F1 and you have now the option to login.
  • If you want this command line permanently available to you, you can set this up in the Virtual Infrastructure client that is connected to your ESX 3i Server..

  • Open the ‘Configuration’ tab of your server
  • Click on ‘Advanced Settings’
  • Open the VMkernel variable and then the Boot list.
  • Scroll down to the VMkernel.Boot.Shell option and enable the select box
  • Now everytime you reboot your 3i server, the shell will be activated. On extra nice option directly below the shell option is the VMkernel.Boot.smallFontForTTY, this will change the shell from 80×25 characters to a 50 line font, which will allow you to much easier work on the shell.

    Customizing the ESX 3i Main screen
    Checking the advanced options list, I found one other nice option called Annotations. Here you can specify the Annotations.WelcomeMessage. What ever message you type here will be displayed on the ESX 3i main screen (when not logged in), so you can specify who is responsible for this server or something creative like that.

    Hacking the ESX 3i Server Root Password
    In case you every are called in to manage an ESX 3i server, but nobody know the root password, here an easy way to ‘fix’ this challange.

  • Again when the ESX 3i server boots and is on the main boot loader screen, hit shift-O
  • In the advanced option line type in: passwdReset="TRUE"
  • Continue the boot process.
  • The root password is now reset and you get full access. Always handy trick to know 🙂

    Well that is for now, if you find more cool tricks let me know and I will add them to the list.

    Enjoy playing with your ESX 3i server.

    VMware introduces ESX 3i

    ESX 3i

    One of the cool things VMware showed this week at VMworld is ESX 3i. ESX 3i is the version of ESX that only run on the VMkernel without the service console (the big linux component). This makes ESX 3i a real single purpose OS only (to run virtual machines), reducing its foot print and increasing reliability and security.

    ESX 3i will be integrated in server hardware coming from HP, Dell, IBM, FSC, NEC and more. The servers do not need local storage anymore as they boot from internal flash sticks.

    The real cool thing I like about ESX 3i is that it has support for SATA disks!! meaning it runs on your notebook 🙂 Jippie!!!! I have successfully tested ESX 3i on my older Dell laptop and it works awesome on my IBM x60. The IBM has a really good SATA controller, that by default is not even supported by Windows XP, but ESX 3i has no problem what so ever with it 🙂 Also the notebook NICs seem no problem for ESX 3i.

    VMware has given all attendees at VMworld a memory stick with ESX 3i on it, so they all can have good fun with it. If you are going to try it yourself, be carefull when you start creating VMFS partitions that you do not accidentally delete your windows partitions.

    Confirmed: Citrix buys XenSource

    Today Citrix has made the announcement that they have come to a definite agreement to buy XenSource for around $500 million in cash and stocks. The actual deal will be done in Q4 this year.

    The original Citrix Press release