ESXi 5 Tips & Tricks: Disable Tech Support Mode Warnings

ESX is officially gone and VMware has made the succession to ESXi complete. Still, some of us might want to log into ESXi's shell by enabling Tech Support Mode or Remote Tech Support via SSH. Now by enabling one of these two support modes, a yellow warning will appear on the host object in vCenter. If you select the Summary tab of the actual host itself, you will see a couple of messages that are immediately apparent:

  • ESXi Shell for the host has been enabled
  • SSH for the host has been enabled

In VMware's defense, it most certainly didn't enable those warning just to annoy us. Enabling Shell and SSH remote access should be used for a limited time for technical support purposes. Security best practices dictate that you keep these disabled--hence, the warning messages. Administering or managing your ESXi host from the command line should be done using the vSphere Management Assistant appliance (vMA) or by installing PowerShell.

That being said, some of us understand the security risks and still want to enable shell access or remote access via SSH, but do not want to see the annoying warning messages that are there just to disturb our perfectly harmonious and well-oiled vSphere infrastructures. To those vNerds I give you three ways of disabling these warnings:

  1. Command line:
    esxcli system settings advanced set -o /UserVars/ESXiShellTimeOut -i 1
  2. Scripted installs you set:
    /adv/UserVars/SuppressShellWarning = "1"

The third method is through the GUI, as follows:

  1. Select your ESXi host and go to the Configuration tab (yes, it has to be done on every host, so use scripted installs to set this if you will deploy on all hosts as a standard)
  2. Click on the Advanced Settings node on the left
  3. Select the UserVars node on the left
  4. Locate "UserVars.SuppressShellWarning" and set the value to 1. (By default, it is set to 0.)

You can disable the warning messages without rebooting the ESXi host or placing it in maintenance mode. For more information about this, check out VMware KB 2003637.

Posted by Elias Khnaser on 09/15/2011 at 12:49 PM3 comments

vSphere 5 Top 10: Storage DRS at #1

The countdown concludes with the top 2 finalist features of vSphere 5. At number 1, I have Storage DRS, a fantastic new feature that load balances virtual machines across the different datastores to optimize performance.

Up until vSphere 5, DRS was a cluster function limited to load balancing compute resources (CPU and memory), and that is a great feature because as VMs start to consume a lot of physical resources on a particular ESXi host, DRS would instruct vMotion to migrate VMs to a more suitable host that can better satisfy their resource needs. In the beginning this was enough--we were just virtualizing some servers and some of us were testing and using it in development environments.

Now that virtualization has gone mainstream in production and the consolidation ratios have increased, it has become increasingly more difficult to load balance VMs across datastores. Moreover, those of us that were successful in an initial deployment of properly load balancing VMs across datastores recognized quickly that maintaining this load balancing was not easy.

Storage DRS now increases the functionality of our clusters by extending VM load balancing across the different datastore leveraging Storage vMotion. As a result of this immensely helpful feature, it's number 1 on my vSphere 5 list.

By now you might have guessed the number 2 feature is the completely re-architected High Availability. Cosmetically HA looks and feels the same to the virtualization admin, but there have been significant changes under the hood.

First, HA no longer requires DNS to properly function. In earlier versions, this was a gotcha that many neglected and it indeed caused issues. That being said, my favorite new enhancement to HA has got to be the fact that it no longer relies on just the management network to determine if a host has failed. The new HA in vSphere 5 will check the management network. If it detects a failure, it will then use the storage connection as a secondary way of verifying if the host really failed. If it cannot contact the host using either method, only then will it deem the host as failed and HA will kick in.

There are other changes to the way HA works, like the fact there no longer is a primary/secondary relationship between the hosts in the cluster. Instead, there's a single master server which is determined via an election process. All other hosts in the cluster are slave hosts. Elections occur at different intervals, but it occurs primarily when a new host is added or if a host is rebooted. There are other things that can also trigger an election but these two events are in the forefront.

Have you been testing SDRS and the new HA? If so, do you agree that these are the two best features of vSphere 5?

Posted by Elias Khnaser on 09/08/2011 at 12:49 PM1 comments

Best Of VMworld 2011: 6 Innovative Vendors

The exhibit hall at last week's VMworld 2011 was the largest I have ever seen. It's amazing how large of an ecosystem VMware in particular has and virtualization in general. It seems like every vendor that had a big announcement saved it for VMworld. For this week's blog, during the show I scouted the exhibit floor looking for what I thought were the best vendors in regards to innovation and I came back with the following list (in no particular order). My Best of Show Awards" go to:

Splunk (don't ask me where the name came from or what it stands for) is a tool that allows you to converge all logs across all your systems in a centralized, indexed and searchable medium. Centralizing your logs and indexing them allows you to quickly search, identify trends, and do proactive monitoring and troubleshooting.

Gale Technologies is a converged infrastructure orchestration and automation tool. I have been looking for an end-to-end tool that can take existing infrastructure of different makes and models and enable orchestration and automation. Gale technologies brings the best of both worlds, providing tools for the Flexpods of the world where the system is pre-built and pre-tested. But it also allows customers that don't have the pre-built and pre-validated infrastructure to also take advantage of orchestration and automation. Gale allows you to take compute, storage and network resources orchestrate ad automate all the way into virtual machines. Really cool stuff!

Nutanix is purpose-built for desktop virtualization, VDI in particular. For a while now I have been saying local disk is dead; don't pitch local disk to me for VDI, I want all the enterprise features of virtualization that require shared storage. Nutanix must have been listening, what they have is a server with traditional compute and local storage in the form of a FusionIO card. The cool thing about Nutanix is they have developed software that allows you to grid their servers together. That way, you get the scalability you need with the shared storage needed at a lower cost. You also get the shared storage needed for the enterprise virtualization features. It's a win-win in a very elegant setup.

Cloupia is another orchestration and automation software that was very impressive. Like Gale technologies they also have software that is dedicated to managing the NetApp FlexPod, but you can also automate and orchestrate existing infrastructure. A big plus and a building block for internal couds.

Tintri is one of the coolest storage solutions for VMware vSphere; it is an NFS solution with SSD drives. The cool thing about Tintri is that it provisions datastores out of the box, which means it presents a datastore to vSphere. However, that is not what caught my attention. What I found impressive were the monitoring capabilities. Today, in traditional storage, if there is a LUN that is being hammered, you only have visibility to recognize that the LUN is being hammered. With Tintri, you can see the LUN, drill into the VM that is causing the contention, and drill even deeper into the VMDK and see which file exactly is causing the issue. For instance, if the windows pagefile is consuming a lot of IOPS, you can drill all the way into the VMDK and see that the pagefile is consuming all the IOPS and then take corrective measures. Tintri represents a new breed of storage that is virtualization-aware or hypervisor-aware. I am very excited about the potential for Tintri.

LG had a very cool announcement at VMworld. They are the first company to release mobile phones that support VMware's Horizon Mobile (previously Mobile Virtualization Platform). The phone is based on Android and delivers a Type-2 hypervisor on the phone, therefore allowing you to run a second operating system with applications completely isolated from your underlying OS. Think of it as VMware Workstation for Mobile Phones. Now you can have your work life and your personal life completely separated on the same device and in the event that you need to wipe out your work profile for whatever reason, you can do that non-intrusively without wiping out the entire phone. It is a very cool life-work separation.

While I am sure all vendors that were present bring innovation and quality products, the vendors mentioned above were the ones that caught my attention. I narrowed this list down from about 30 vendors that I was able to investigate.

Disclaimer: I do not receive any direct or indirect compensation from the listed vendors. I looked at them from a purely independent stand point.

Posted by Elias Khnaser on 09/06/2011 at 12:49 PM6 comments

Top 6 VMworld Takeaways

VMworld 2011 in Las Vegas can only be summarized this way: It's The Godfather of all IT conferences. With an impressive 20 thousand-plus attendees (and there may have been more, had Hurricane Irene not interfered).

The dominant subject at VMworld was vSphere 5 and the products that hover around it, from SRM to vCloud Director, vCOPs and others. By the time I got to Las Vegas, I was honestly already ready for vSphere 6. The reason for that is, I had been playing with the vSphere 5 beta for months, then started blogging about it in detail here on and then David Davis and I did a marathon one-month recording of the new TrainSignal VMware vSphere 5 training (shameless plug).

Now, hopefully you have been following my feature countdown and you know I still owe you all one last blog with my top two features. We'll hold off for next time. Today, I want to tell you briefly about some of the end-user computing announcements made at VMworld and in the coming weeks, I will offer some details on them. By order of importance, here are the announcements:

#1. VMware Mobile Virtualization Platform (MVP) has now been renamed Horizon Mobile and this is by far the best announcement, a hypervisor for mobile phones. Today, it only supports Android and allows IT to deploy a Type-2 Hypervsor to an existing Android phone that is preconfigured with all the enterprise applications, settings and features. Furthermore, it allows IT to now monitor and meter how much the user is using the enterprise VM and can now build expenses and allowances based on how much the enterprise VM was used from a data perspective. This will change everything.

#2. Project Octopus is AWESOME! A Dropbox-like solution for the enterprise, the idea here is to have your files synced and accessible from any device, anywhere. The difference here is, instead of having it unmanaged, Octopus will allow IT to set policies, run reports about which files were uploaded, downloaded, where you can and cannot access this data from, etc. What we have here is a decent balance between the consumerization of IT-style services that we have grown to love and the IT policies that are needed to protect the business. Since IT will be transparent in its enforcement, this is a huge win-win.

#3. Project Horizon is being touted as a universal broker that can connect VMware View, Horizon Mobile and Project Octopus. The idea here is now we have all these solutions, so how do we bridge the gap between them, how do we connect them so they make sense, integrate and work together rather than being separate islands? The answer is the universal broker that would allow the same user logging into View to get their files and would allow that same user to also get their files on their mobile devices using Octopus.

#4. ThinApp Factory will automate software packaging. Think of it as a P2V except for applications: Instead of sequencing, ThinApp Factory will analyze the application, then will convert it into a ThinApp-virtualized application. This is huge because it is simplifying the process. Hey Microsoft, can you please wake the hell up and innovate just a bit? You have had Softgrid, I mean App-V for how many years now? Why, oh why, did this technology not come from you?

#5. ThinApp Factory is also supposed to connect to or convert Citrix XenApp, Remote Desktop Services (RDS), and App-V. I think they meant View will now support connectivity to these systems but I will verify that.

#6. AppBlast, which will allow the ability to deliver applications remotely and seamlessly using any browser, was the WOW demo in my opinion. AppBlast will be based on HTML 5 and what they showed us looked very similar to what you would see from a Citrix XenApp application. The difference: AppBlast will be embedded in a browser, with no plug-in needed. Now what I don't know yet is, where is this application coming from? Where is it hosted? What remote protocol is being used, if any? How does it scale? Those and about a hundred other questions I'll be try to get answer to from VMware in the coming weeks and months.

That pretty much sums it up for me from an end-user computing stand point. I will elaborate more on all these points as I gather more information.

In the next few weeks (or sooner), I will also be blogging about the Best of VMworld: Vendors, a list of the most interesting vendors I saw out there with their cool technologies. In the meantime I would love to get your feedback/comments on this blog, so share 'em!

Posted by Elias Khnaser on 09/01/2011 at 12:49 PM1 comments

My VMworld 2011 Session Recommendations

We'll take a break from my countdown to get to something a bit more time-sensitive: We are less than one week away from VMworld 2011 in Las Vegas! With about 20,000+ attendees expected this year, it will, without a doubt, be a fantastic show that I am looking forward to. For those attending, rumor has that VMware has bought the entire stock of Blackberry tablets, aka PlayBook--at least that was the rumor on Twitter.

This year, Rick Vanover, David Davis and I have decided to start what we hope will become a yearly tradition of recommended VMworld sessions. While we are sure all the sessions will be fantastic, there is no way for us to attend all of them. Because of the large number of sessions, it could be overwhelming for some to choose sessions to attend, so we decided to help.

Now before I give you my session recommendations, it is worth noting that this year VMware requires you to pre-register for your sessions. You no longer can just walk up and scan your badge, so I recommend that you go to, login and use the Schedule Builder to register for your sessions to stay organized.

Alright, let's get down to the session selections. Here are my picks: the ones in bold are ones I will be attending; the others are ones I highly recommend:

  • CIM1436 -- vSEL: Building VMware's Hybrid Cloud
  • SEC2114 -- Customer Panel: Ensuring Compliance In a Virtual World
  • VSP2347 -- What's New With vSphere 5
  • CIM2561 -- Stuck Between Stations: From Traditional Datacenter to Internal Cloud
  • EUC4080 -- Integrate "Social" Into The Enterprise To Improve Productivity
  • CAP1707 -- 7 Steps to enterprise PaaS
  • CAP2471 -- A Customer Scenario for Next-Generation Data Management with vFabric
  • VSP1682 -- vSphere Clustering Q&A
  • VSP1956 -- The ESXi Quiz Show
  • EUC2956 -- VMware MVP - Solving Enterprise Mobility Challenges
  • BCA1360 -- How A Global Enterprise virtualized Exchange 2010
  • VSP2447 -- Understanding Virtualization Memory Management Concepts
  • SQL Server on VMware ESX BCA1995 -- Design, Deploy, Optimized
  • BCA1931 -- Design, Deploy and Optimize SharePoint 2010 on vSphere
  • CIM2449 -- Introduction to vCenter Operations
  • CIM2452 -- vCenter Operations Technical Deep Dive
  • VSP1425 -- Ask the expert bloggers
  • EUC2103 -- Storage tips and Best Practices for View Deployments
  • BCA1985 -- Oracle Database best Practices on vSphere
  • BCO2479 -- Understanding vSphere Stretched Clusters
  • VSP1999 -- ESXTOP for Advanced users
  • BCO1946 -- Making vCenter Server Highly Available
  • BCA1230 -- Enterprise Java Applications and Virtualization Best Practices
  • EUC2692 -- Rethinking Storage for Virtual Desktops
  • EUC2846 -- View Enterprise Architecture Design and Implementation Best Practices
  • BCO3336 -- DR To The Cloud - Service Provider Perspective

You don't want to miss VMware Mobile Virtualization Platform - MVP; find the labs, find the demos and see how it works. Orchestration and automation are key for private clouds. Plenty of vendors have products around those topics this year, so make sure you check them out.

If you are attending the show, drop me a note on twitter @ekhnaser. I'd love to meet you so you can tell me how much you agree/disagree with my blogs!

Posted by Elias Khnaser on 08/23/2011 at 12:49 PM4 comments

vSphere 5 Top 10: VMFS-5 at #3

If you've been reading this blog long enough, you know by now that I'm a big fan of VMFS. So, it should come as no surprise that VMFS-5 landed in the No. 3 spot on this countdown.

VMFS is a purpose-built file system optimized to run virtual machines. One of the big drawbacks of VMFS-3 was its size limitation of 2TB. With the introduction of the vSphere APIs for Array Integration, it was simply a matter of time before that limitation was lifted. VAAI has several useful features, but one in particular is more relevant to our current discussion: VAAI offloads SCSI reservation from the hypervisor onto the storage array, thereby significantly enhancing the performance of the VM in general.

VMFS-5 now supports datastores of up to 64TB in size, thereby stripping NFS from its only strategic advantage as far as I am concerned. The other feature enhancement that is now enabled within VMFS-5 is the ability to leverage VAAI with thin provisioning and deliver automatic free space reclamation. This is huge. Prior to this version of VMFS, many did not know that when you use thin provisioning and as you start deleting data, you need to go back and reclaim that free space. Well, the problem was that it had to be done manually. And so for those who did know about the free space that could be reclaimed probably had some manual methodologies or scripts that reclaimed the free space. Now, you will be happy to know that another annoying task has been removed from your to-do list and has been automated.

Now, I am sure the last two features in the countdown are probably going to be easy to guess. So, this time a new question: Between Storage DRS and the new High Availability, which one is your top feature and why?

Posted by Elias Khnaser on 08/18/2011 at 12:49 PM5 comments

vSphere 5 Top 10 Countdown: vCloud Director 1.5 at #4

We are getting down to the wire with our Top 10 countdown, with three features left in the spotlight. In this spot I've chosen vCloud Director 1.5. While this is not necessarily a feature of vSphere 5, it relies heavily on some of the new vSphere 5, not to mention that VMware made it part of the announcement a few weeks ago.

vCD has three features that I like a lot. First off is, of course, Linked Clones. Virtualization introduced us to fast provisioning of VMs, but it was at the expense of cloning the entire disk, resulting in the unnecessary duplication of hundreds of Windows operating systems. Linked Clones, as with VMware Workstation and VMware View, allow you to provision VMs faster by not necessarily performing a full clone, but rather what's called a "fast clone." This means you still have reliance on the master image, but you commit your changes to a difference file. This saves significantly on storage costs, de-duplication software and more. Something tells me that in the next update of vSphere 5, maybe update 1, we will see Linked Clones in vSphere as well, as vSphere seems to be the only tool that lacks this Linked Clones functionality.

I also really like is vCD Messages. In all of my deployments, executive IT always wants some type of visibility into the environment. They want to know by simply looking at a console or a mintor that something is going on in the system, and they want to know when it is fixed. vCD Messages is a perfect companion in this scenario. It will tie into your existing CMDB or ticketing system and can display notifications and messages. vCD comes bundled with up to 100 tasks out of the box -- pretty cool!

And finally, I also like that vCD has good integration with vShield. It's a major enhancement and now supports:

  • Five-Tuple Firewall services, which means you can now build your firewalls based on destination IP, destination port, protocol, source IP and source port
  • IPSec VPN Services allow the user to establish site-to-site VPN tunnels. This is pretty powerful and unique to vCD in that as a customer or user of vCD you can create this site-to-site yourself. It's completely self-service and pretty cool.

To wrap it all up, vCD 1.5 now has support for Microsoft SQL Sserver, which means you don't need an Oracle license anymore to deploy this. SQL Server support is a major enhancement, as most of my customers were hesitant about vCD because they did not want to introduce Oracle and Oracle licensing into the environment.

What do you think of vCD 1.5 and does it deserve such a high ranking in my Top 10 list? Post your comments here.

Posted by Elias Khnaser on 08/16/2011 at 12:49 PM8 comments

Citrix Chips Away at Another Desktop Virtualization Layer

Citrix takes another full swing at VMware with the acquisition of RingCube, just a few months after the acquisition of Kaviza. Citrix is sending a strong message that when it comes to desktop virtualization, they are not settling for 1st place, they want to make sure that the closets competitor to XenDesktop is in 5th place.

Now, I had predicted that VMware would make a move and acquire Unidesk to significantly reinforce VMware View, but that has not happened nor has there been any indication VMware is even interested in a desktop virtualization acquisition. Nonetheless, I still think Unidesk is a desperately needed addition to View.

Now that being said, let's examine the timing of the announcement. Citrix finally learned how to make announcements that can make an impact. Last time, it announced the acquisition of Kaviza. VMware followed with the announcement of vSphere 5 and that completely overshadowed everyone else's big news. This time, Citrix timed it right by announcing RingCube weeks before VMworld, but also weeks before VMware announces View 5 which will finally support a profile management solution from their acquisition of RTOSoft a few years back.

The message Citrix is sending is, while View 5 finally supports profile management, XenDesktop has had that for years but now XenDesktop takes leaps and bounds forward with the support of user-installed applications, thereby solidifying XenDesktop even more.

RingCube, for those of you that don't know, enables users to install applications to their own desktop. It combines the best of both worlds: Users still get a non-persistent desktop that is locked down with IT-supported applications and settings, but it allows the users to install their own applications into a separate virtual disk, a separate layer which is merged with the IT-delivered layer seamlessly. From a user perspective, it is completely seamless; from an IT perspective, it is very manageable. It's Win-Win.

I could not tell you how many use cases I have come across where customers are interested in desktop virtualization, understand and see its value but cannot accept the persistent or non-persistent approach, and profile management alone was never enough. User-installed application support is an imperative feature that is most critical to some customers and certainly makes the migration from physical machines to a virtual environment that much easier. At the same time, it lowers storage dependency and cost.

In summary, Citrix is making all the right moves when it comes to desktop virtualization, and it has turned XenDesktop into a beast that addresses every aspect of the technology. Check this out:

  1. XenApp for published and streamed content
  2. XenDesktop for VDI and hosted applications, Kaviza solidifies this option and simplifies it
  3. Provisioning Services for streaming to physical and virtual machines
  4. XenClient for Type-1 Client Hypervisors/off-line use cases/BYOD
  5. RingCube for user installed applications layering
  6. NetScaler for remote secure access
  7. Branch Repeater for WAN acceleration of ICA /HDX

It is hard to compete with this list, yet I still think VMware can compete with a few acquisitions and some product releases like the ever-awaited Client Virtualization platform which I intend on following on with Stephen Herrod this year--again--at VMworld.

VMware also has the Mobile Virtualization Platform, which Citrix does not have yet. That's a huge plus in VMware's corner. I will make sure that is the topic of another blog. In the meantime, I really want to hear your analysis on this. Let's hear it!

Posted by Elias Khnaser on 08/12/2011 at 12:49 PM3 comments

vSphere 5 Top 10 Countdown: Swap to SSD and VAAI at #5

The countdown to the number 1 VMware vSphere 5 feature continues, and this time we look at Swap to SSD and VAAI, which are tied in the number 5 slot.

The Swap to SSD feature is of particular interest to me because of its overcommitment potential, especially as it relates to desktop virtualization. This cool new feature allows the VMkernel to automatically detect and tag any SSD storage that the ESXi host has access to, whether it is local storage or network storage. Once detected and tagged, the VMkernel will then instruct the scheduler to take advantage of this storage by allowing the ESXi swap to extend to it.

While I am not suggesting by any means that you use SSD as a replacement for memory (and it still is best practice to add physical RAM if you are swapping), this cool new feature does let you take advantage of SSDs in the amplification of memory overcommitment.

As for vSphere Storage APIs for Array Integration, or VAAI, it has always been one of my favorite storage features. With vSphere 5, it got better. One new capability I am particularly excited about is the automated free disk reclamation for thin provisioned disks.

Prior to vSphere 5, the process of reclaiming free disk space as a result of using thin provisioning was a relatively manual process. I am sure some of us have automated and scripted it, but I still consider this a manual process--you still have to maintain the script. Another capability of VAAI I am excited about is that it now extends support for NFS; it was previously just supported on Fibre Channel.

And yet another cool capability of VAAI is known as NFS Space Reservation, which gives you the capability of reserving space on NFS. With NFS, one of the selling points is that it is thin-on-thin by design--the datastore grows and shrinks automatically. There are times I wished I could reserve all the underlying space and not necessarily have to do thin-on-thin. Now you can accomplish that with the new VAAI.

The last feature of the new VAAI I want to share with you is also around NFS and known as Full Copy, which can offload certain copy operations like "clone" or "deploy from template" from the host to the array. Now before you flame me in the comments, I understand that certain NAS vendors had vCenter plugins that would allow you to do this from within vCenter. The enhancement with vSphere 5 is now you can do it natively using the traditional vCenter commands, provided that your storage vendor can support it, of course.

So, guess what I've got at #4. That's for next time.

Posted by Elias Khnaser on 08/09/2011 at 12:49 PM0 comments

VMware Listens, Changes vSphere 5 Licensing

It is not often that a technology company truly listens to its customers and openly corrects its position. Usually, changes are made months and years later and the change is introduced in a subtle and stealthy way.

VMware today bravely changed the licensing and corrected it. I'm ecstatic at the change, but I'm even more excited to hear that there's a culture at VMware that takes its customers seriously, one that is saying "we listen, we understand and we are the customer's champion."

For the record, as soon as VMware announced officially the vSphere 5 licensing, I wrote on Twitter that VMware cannot sustain this licensing model and that it would be changed within 60 days. I also wrote on my hope that Stephen Herrod announces the change at VMworld. I am happy that this change came earlier rather than later. I had feared it would backfire on vSphere 5 had this not been amended.

You can see the new changes here, here and here, so I wont' repeat them in this blog. The new limits are significantly better and completely fair and workable. The changes make the upgrade easy going from vSphere 4 to 5. Keep in mind that this is still a per-processor license, so if you have a dual-socket server, you will need two licenses; if you were purchasing the Enterprise Edition license, that means you are entitled to 128GB vRam.

Now keep in mind that the vRAM is cluster-wide: If you have four or five ESXi hosts in a single cluster, the vRAM entitlement is pooled between all the hosts in that cluster.

One other thing to pay attention to is that in any vSphere environment, you have to take into account high availability. In this case, if you plan on dedicating or assigning 128GB of memory to your VMs, you will need to have at east 20 to 25 percent extra physical memory for HA. This physical memory does not count against your vRAM entitlement.

Let's take a real world scenario. If you have a server with two sockets and you plan on provisioning 32 VMs with 4GB of memory, then your vRAM requirements would be 128GB. So, you can get away with purchasing the Enterprise Edition. However, while you are licensing the server for 128GB of vRAM, you should have 25 percent more physical memory in that server for HA. The configuration of your server would end up being 160GB of physical memory.

So, what about VDI? VMware thought about that too, and gives us the vSphere Desktop Edition, a licensing package that is dedicated to virtual desktops. This edition has no limitation on vRAM, but has a limitation on the number of powered on VMs. The limit is 100 VMs per ESXi host. Again, that is a very fair and acceptable ratio; most VDI environments have 65 to 75 VMs to host, so it is perfect.

The Desktop Edition is also not limited to VMware View, so for those that are deploying Citrix XenDesktop on vSphere, this edition is a perfect compliment.

Finally, I am very proud to be one of the many that voiced their concern about vSphere 5 licensing and I am also very proud to be part of the VMware vExpert community and congratulate its major contributors on making sure that our feedback got to the right place to make a difference. This type of culture will ensure VMware's continued success.

Posted by Elias Khnaser on 08/03/2011 at 12:49 PM5 comments

vSphere 5 Top 10 Countdown: ESXi Firewall, vSphere Web Client and Storage vMotion at #6

Six is a party and at number six are three features I put at a tie: ESXi Firewall, vSphere Web Client and Storage vMotion.

Let's start with Storage vMotion. Up until VMware vSphere 5 you were unable to use Storage vMotion if the VM had a snapshot attached to it. With vSphere 5, that limitation will be lifted thereby allowing a storage vMotion operation to successfully complete . This makes a lot of sense as an enhancement in vSphere 5. Without it, vSphere 5 wouldn't be able ot implement one of the most anticipated features: Storage DRS.

Also tied at number six is the new and enhanced vSphere Web client. In older vSphere versions, we had access to a Web-based administration interface to vCenter and the ESX/ESXi host. That interface, though, had limited functionality. vSphere 5 extends the reach of the vSphere Web client, allowing you to perform the vast majority of tasks that you could perform from the regular vSphere client today. My only disappointment here is that the vSphere 5 Web client is not HTML5-compliant.

The third in this trinity-featured countdown goes to the new ESXi Firewall. It's a new service-oriented, stateless firewall used to protect the management interface of ESXi. It also eliminates the reliance on IP tables and rule sets to define port rules for services. The firewall can be configured either from a GUI through the vSphere client or via the command-line using esxcli.

As the countdown to the number one vSphere 5 feature continues, which features do you think will land in the top 3 spots, and in which order?

Posted by Elias Khnaser on 08/02/2011 at 12:49 PM0 comments

vSphere 5 Top 10 Countdown: VM Capabilities at #7

The countdown to the number 1 feature of vSphere 5 continues. The number 7 spot goes to vSphere's virtual machine capabilities. vSphere 5 is a platform upgrade and naturally there will be feature enhancements to its VM capabilities. After all, this entire environment is built to host VMs. So what's new? Plenty, take a look:

VM Hardware version 8 unlocks the capabilities of VMs running on the vSphere 5 platform. Upgrading your existing VMs' virtual hardware gives you access to the following capabilities:

  • Up to 32 vCPUs. Do you think you can tap all this vCPU power? Any excuse that applications need CPU processing power?
  • Up to 1 million IOPS. That's a lot of IOPS for a VM to handle; are there any apps that can't run with that?
  • Up to 1TB of virtual memory. I still don't see physical servers with 1TB of memory, but if there's ever a need, now VMs can handle that load as well.
  • Support for client connected USB. While I welcome this feature, it is limited to USB devices that are connected to the machine on which you are using the vSphere client or web client. So, you still cannot connect a USB in the ESXi server and pass it through.
  • Support for USB 3.0.
  • Non-hardware accelerated 3D capabilities. It's a great enhancement for Windows Aero and VDI in general.
  • Smart Card Readers. Similar to client connected USB devices, I welcome the addition of support for Smart Card readers. Again, it is limited to the host you are using the vSphere client and web client from.

Some other useful features:

  • GUI for multicore vCPU configuration. It's very similar to what you see in VMware Workstation where you can assign the number of vCPUs and vCPU cores.
  • Apple Mac OS X Server guest operating system support. vSphere 5 now supports Mac OS X 10.6 Snow Leopard as a guest OS in a VM, albeit it is only supported on Apple Xserve3.1. Nonetheless, this is a very welcome step in the right direction.
  • VMware Tools versions support matrix. New version of ESXi means a new version of VMware Tools. The cool thing here is that VMware Tools from version 4.x are supported on 5.x and vice versa, so you can have VMware Tools running version 5.x on ESX/ESXi hosts running vSphere 4.x

It is quite an impressive collection of capabilities that you can now assign to a virtual machine. Looking back, I would never have thought we would come such a long way with empowering VMs in such a short period of time. The excuse against virtualizing certain types of applications and servers is becoming very slim, one would truly have to fight hard to make the case for a physical server after deploying vSphere 5.

Posted by Elias Khnaser on 07/28/2011 at 12:49 PM3 comments

Subscribe on YouTube