The story of a difficult experience with Hyper-V
If you follow my Twitter feed you knew this article was coming…
After spending about 3 hours on the phone with a customer trying to help them get their CSV (Cluster Shared Volume) for Hyper-V back online (I was on the phone not because I’m a Hyper-V guy but because I implemented the storage they use for Hyper-V) I decided that maybe I should do a deep dive and learn a little more about how CSV’s work so that I can better compare them to VMFS. VMFS is the clustered file system that VMware uses to share SAN LUNs between physical VMware servers. This article also goes on to explain why Hyper-V is an inferior solution due to some of the other issues that I’ve seen in my limited experience with Hyper-V.
Anyhow, before we get to the technical stuff I need to point out that I am obviously biased to VMware… it’s what I do every day. I will try to be as fair as possible, but lets face it if you are just looking for the take away without reading the whole post it is that VMFS is far superior to CSV’s, and that while VMware might look more expensive on a bill of materials it will probably save you time and money in the long run.
The Technical Stuff
Cluster Shared Volumes have been around a long time, Microsoft has been using them for everything from Exchange clusters to SQL clusters. They adapted CSV’s to work with HyperV so that they could allow virtual machines to more easily move from one HyperV host to another, similar to what VMFS allows VMware ESXi servers to do. Both enable high availability for virtual machines because if a host fails, other hosts can access the virtual machines the failed host was running. Additionally CSV’s are needed because under them is NTFS… which was never designed to be accessed by multiple systems at the same time, because of this something had to be put in place to allow that to happen.
Ok so the first article I come across on Technet has this to say:
“…the Cluster Shared Volumes feature included in failover clustering is only supported for use with the Hyper-V server role. The creation, reproduction, and storage of files on Cluster Shared Volumes that were not created for the Hyper-V role, including any user or application data stored under the ClusterStorage folder of the system drive on every node, are not supported and may result in unpredictable behavior, including data corruption or data loss on these shared volumes. Only files that are created for the Hyper-V role can be stored on Cluster Shared Volumes. An example of a file type that is created for the Hyper-V role is a Virtual Hard Disk (VHD) file. Before installing any software utility that might access files stored on Cluster Shared Volumes (for example, an antivirus or backup solution), review the documentation or check with the vendor to verify that the application or utility is compatible with Cluster Shared Volumes.” Taken from: http://technet.microsoft.com/en-us/library/dd630633%28v=ws.10%29.aspxSo to me, that means that CSV’s are flaky, to say the least… but let’s continue.
VMFS, on the other hand, can store pretty much anything you can upload to it… zip files, iso files, etc etc… VMFS is almost like LVM in linux, it doesn’t care what you put on it.
My Next Point
All HyperV nodes that are using a CSV are at the mercy of the coordinator node for that CSV. Think of it this way, you need to look for something that is in a filing cabinet, but before you can actually get the folder you need you must first talk to the secretary and ask her if its ok to look at the folder. In more technical terms this means that the coordinator node keeps track of all the metadata and file locking on folders on the CSV, after the coordinator node allows you access to the folder then IO to things in that folder happen directly to the LUN. But don’t take my word for it… Microsoft explains it in this article http://support.microsoft.com/kb/2008795
What scares me about this method is that if something gets hosed with the Coordinator node and the Failover Manager doesn’t fail over properly your CSV is inaccessible. And going by the reliability track record of Microsoft Services I would not bet my job on Failover Manager 🙂
VMFS, on the other hand, is a clustered file system that has no owner… there is not anyone node that controls access to the file system. File locking is done at a file level by a ‘pulse field’, and in this field, a host must periodically update its time stamp and let the file system know that it is still using the file. If a host crashes and another host wants to use the file the host can let the file system know that the timestamp hasn’t been updated lately and that it is taking over ownership…. this means that each host can access files in the event of a node failure without waiting on a response from a centralized management node. If you want the in-depth answer check out this article http://blogs.vmware.com/vsphere/2012/05/vmfs-locking-uncovered.html
Next, let’s talk about how to get HyperV to actually work …. and I’m not talking about just getting a VM to boot up. I’m talking about setting up HA and DRS and automatic load balancing etc. With VMware you group all of your physical servers into a Cluster and then check two boxes… one to turn on HA and one to turn on load balancing. Of course, you need to set up network interfaces for vMotion… but other than that your done. Oh, and by the way, this was all done from the VMware vSphere client.
On to HyperV… lets see where do we start…. no really… which interface do you want to start with. If I want to configure virtual machines I’ll need to use HyperV Manager… or maybe Systems Center Virtual Machine Manager (SCVMM). If I want to set up CSV’s to allow for HA to take place I’ll need to fire up MS Failover Manager. On and if I want to actually format a LUN, well then I’ll need to get into Disk Management.
I’m sure you get my point….
We could talk about virtual switches and distributed virtual switches (or the lack there of), or things like NIC teaming and how it is implemented, but I don’t want to write a book.
The take away
OK so normally the number one reason I hear that people are using HyperV is because it comes with Windows … Its free. But is it really free? I could argue that it takes less time to implement VMware than it does HyperV… and isn’t it known that “Time is Money”… it certainly is if you are paying someone to set it up.
I could also argue that 3+ hours of downtime trying to resolve a CSV issue where VM’s are not accessible is definitely a loss of productivity and in turn a loss of money.
Lastly SCVMM is not free and while it is not required, if you want to compare apples to apples you will want it…
So in the end VMware may be an additional line item on a bill of materials, but in the end it may be the best damn investment you will make for your virtual environment.
If you are not convinced yet here are some other fun stories about the HyperV
- http://www.ms4u.info/2011/05/why-you-should-not-running-domain.html <- Failover manager required AD, if you virtualize all your Domain Controllers Failover Manger wont start.
- http://blogs.technet.com/b/chrad/archive/2010/07/15/cluster-shared-volumes-csv-extending-a-volume.aspx <- Expanding a datastore in vmware is as simple as right clickign the lun to extend, selecting expand, clicking next like 3 times…. check out this link for the process for a CSV 🙂
- do a google search for HyperV live migration using multiple NICs…. you wont find much, so if you buy that server with 128 or 256 GB of ram and need to take it out of production for maintenance… better have 10Gbps networks in place or grab dinner and a movie… (1Gbps network can theoretically move 125MB/sec or 7.5GB/minute so to move 256GB that would be about 34 minutes… so if you want to do a rolling outage multiply the number of hosts you have by 34 minutes to move all that ram…. just one more way VMware costs you less)
- Look up how to do nic teaming in HyperV… Then look up how to do it in VMware
The bottom line
If you want a solution that is straight forward and easy to use then HyperV is probably not the way to go. While it may not be a big line item on a bill of materials, what it saves you there it is certainly to cost you in time and effort (not to mention frustration).
Have you ever used server 2012 with hyper-v or even vmm 2012? So many of the things you write are just plain wrong, hence my guess would be you simply have not tried the products? Hyper-v was improved alot from 2008 r2 to 2012.
The customer I write about is using HyperV 2012 and its “improvements”. Oh and I welcome you to enlighten me as to how microsoft has fixed those problems. If they have then I would love to know about them so that i can pass that information on to the customer. However since they had Microsoft on the phone doing support I find it hard to believe that much has changed … unless of course you mean SMB3 shares as your shared storage …. in which case you render all of EMC Storage useless for HyperV customers.
Justin, the thing that worries me about you email is that your sources are out-dated. All your links refer to 2010 or even earlier in some cases and the support.microsoft.com articles mention speficially Hyperv 2 or Windows 2008. More specifically in place if the adding storage to CSV you should reference the following article http://blogs.msdn.com/b/clustering/archive/2012/04/06/10291490.aspx which details the process for Windows 2012. While I wont argue with you the ease at which I can stand up a VMware environment, you might want to be more factual or up to date on your knowledge.
A lot has changed in Windows Server 2012.
Speaking about your 4 last points:
1) It was an issue in Hyper-V 2008 R2. In Hyper-V 2012 it is no issue anymore. (http://technet.microsoft.com/en-us/library/hh831414.aspx)
2) Actually it is as simple as with VMware if you use the Disk Management GUI.
3) Hyper-V 2012 indeed has no multi-nic vMotion capabilities (to the same host) as VMware does. But if you vMotion multiple VM’s to different hosts then you can leverage multiple teamed nics.
4) It was not very easy to team in Hyper-V 2008 R2, but in Hyper-V 2012 it is as easy as in VMware. (http://4sysops.com/archives/nic-teaming/)
Then explain why it happened to a 2012 hyperV cluster it is is fixed ?
If you are using Hyper-V in Server 2012, then your comment about Failover Cluster Manager needing access to AD is not correct. In Server 2008 R2, yes the Cluster did need access to AD, as otherwise the Cluster would fail to start. However, we can now start the Cluster in Server 2012 without needing a DC being online which means that all the Windows DCs can now be virtualised:
http://blogs.technet.com/b/wincat/archive/2012/08/29/windows-server-2012-failover-cluster-enhanced-integration-with-active-directory-ad.aspx
I would actually read into Server 2012 before making such false comments…
Great article Justin! Apparently the MS marketing machine over-promised again with the release of Hyper-V 3 (duh).
Wat about the fact that in 2012 snapshots corrupt VMS, I am at tech ed and the me guy that did BAM s cm corrupted when he tried to roll back?
Not knowing the details, I can’t explain what happened or why.
I know that if a coordinator role goes down then another one is elected with a quorum mechanism. Different CSV volumes also (can) have different coordinators.
That said, I can’t argue that there are (unplanned) ways that a CSV volumes goes down, just like any other clustered file system volume. I have seen VMFS volumes disappear because of APD/PDL.
Bottom line, failures can and do happen in the IT (because of bad code, bad design, or just because of pushing the wrong button).
That’s ok, it keeps us off the streets 😉
Thanks for the link
Alright, good to know they are already on 2012, however as many other have pointed out, most, if not all, your points then become sort of pointless, as they are just plain wrong/FUD.
As to why it failed, the lack of knowledge and/or experience with the software platform would be the most obvious guess.
Yeah… thats why we called MS support… they changed a setting and it took them 3 hours to fix it.
As to why it failed, the lack of knowledge and/or experience with the software platform would be the most obvious guess. <—
what a plain bullshit.
so mentioning it has been problematic in the past is irrelevant because there's a new version that doesn't fall apart as much anymore.
Wow, great, that totally beats good design.
And if the latest greatest version dies away during production use, that you interpret as a lack of knowledge – hello? the key thing is a hypervisor should be reliable and resilient.
If someone with reasonable virtualization background can't make it run mostly reliable right away, then this is not a grand thing.
Blaming the messenger is quite useless.
LOL
I’ve deployed Hyper-V 2012 clusters, tried it again with an open mind as I did with Windows 2008 R2, only to be punished time and time again with 2008 R2. So many problems with 2008 R2. Is 2012 better? So far, yes, it is a significant improvement. Does Justin have good points about Hyper-V in Server 2012? Absolutely. It’s still a monumental PITA to get a cluster up compared to VMware. NIC teaming is easier to setup. However, there’s nothing that says the teaming works better, as in it fails less, which was my major problem with 2008 R2. In about six different customers, most of which this happened multiple times, a 2008 R2 VM would have its entire TCP/IP stack effectively blown up. In every case, the commonalities was NIC teaming was involved, the guess and host were 2008 R2 with Hyper-V. Nobody, including MS support, could ever give an explaination as to why, and it would happen like once every six months within customers who suffered from the problem, so it was really difficult to track down. We tried updating drivers, ensuring switch configs and NIC teaming was setup right, keeping the host and guests patched to current, different vNIC types, you name it, we tried it. In 2012, we’ve already had issues with hosts when enabling VM Queuing, which is recommended by MS, but it’s not worked for some customers.
Other things too that are just ridiculous in comparison. We deploy a lot of Cisco UCS blades, which lend very well to boot from SAN. With VMware ESXi, not much space is needed for these boot LUNs, but Hyper-V? Consider you need enough space for Server 2012 AND for a crash dump in case you need it for troubleshooting. With virtualization hosts now often approaching 128GBs or more of RAM, that’s A LOT of space to waste on the SAN to facilitate a boot volume. Even if you’re not booting from SAN, it’s still impractical to use an SD card as the boot volume for Hyper-V due to needing a big pagefile.
Hyper-V in Server 2012 is a significant improvement, but I personally would run my infrastructure on VMware.
Hi Justin,
Please do your homework first. I cannot take your blog seriously if you pile error upon error, mixing CSV v2 and CSV v3 and quoting outdated references. Check the excellent and free videos on CSV in http://channel9.msdn.com/Events/TechEd/NorthAmerica/2012/WSV430.
I’m sure I can mess up a VMware cluster without proper knowledge of how to configure storage, servers and network for Hyper-V in Windows Server 2012 including the required updates and hotfixes.
So again, don’t make yourself a laughingstock.
Regards,
Hans Vredevoort
MVP Hyper-V
V2 or V3… they still broke…..
but since you seem to be the hyperV expert how about you put a post together with EVERYTHING you need to run hyperV with all its options…. and ill do the same for VMware…. for say 6 hosts (dual Socket system 256GB ram each) and we will compare the prices ?
Wow there are a lot of things wrong in your blog post. I think I can find a mistake or a misunderstanding in every argument. I was you I would do a little more research next time 🙂
btw. I would trust CSV a lot more than VMFS
HA, DRS, NIC teaming and even distributed vSwitch (in Hyper-V world it is called logical switches) are all possible within SCVMM 2012 SP1 in combination with Hyper-V R3. You can easily set up an entire Hyper-V cluster using only one console, SCVMM console, as with vCenter in VMware world. Only thing needed is a storage solution capable of talking to SCVMM with SMI-S or SMP protocol (as most recent storage solutions do btw).
So, as said by others in the comments, please do read a little bit more about SCVMM 2012 SP1 capabilities. As a matter of fact, setting up the logical vSwith in SCVMM using multiple nics (setting up a NIC teaming config for your vSwitch) gives pretty much the same experience as what I did in the past using vCenter. Same way of thinking is used, even in the GUI.
It is indeed necessary to use SCVMM when you want the ‘single pane of glass’ for setting up and maintaining your Hyper-V clusters, but isn’t that the same story with VMware and their vCenter? From my experience in the past with VMware, it is I think.
And about the corrupt CSV: not so long ago, we had the exact same problem with VMFS volumes. VM config files which were corrupted when a power failure occurred at one of our customers. So don’t pretend that this can’t happen with this file system, because it really CAN get corrupted, as stated by Hans Vredevoort.
Regards
Stijn Verhoeven
Consultant Virtualization Solutions
1.) show me pricing for this single pane of glass ? do i really save anything over VMware ? (yet again no one wants to pony up pricing for the Cadillac solution lol)
2.) i never said that VMFS cannot have corruption…. however was the volume still mountable by the ESXi hosts ? Sounds like it if you could see that a VMX was corrupt … so what im getting at is that the “CLustered file system” design in VMware is BETTER than CSV’s… i was not talking about the reliability of VMFS over NTFS…. which im guessing if we were to gather statistics on VMFS failures vs NTFS failures we would find that NTFS gets corrupt more than VMFS
I would love to see comparative post as well as comparative pricing ….
Its interesting… you post some stuff about hyperV on a known vmware blog and my servers get hammered LOL
Last year I wrote a series of blogpostings in which I compared vSphere 5.1 with Windows Server 2012 Hyper-V. Have a look here:
http://up2v.nl/2012/09/06/vsphere-5-1-versus-windows-server-2012-hyper-v-replication/
You will be amazed by the number of improvements in Server 2012 Hyper-V. Server 2008 Hyper-V had quite some issues and you found a few of those ;-). I guess all of them are gone in 2012.
A cost compare is very difficult. Quite soon you will be comparing oranges to apples. System Center 2012 delivers a lot of features. VMware solutions as well. But different features.
Hello all, interesting thread!
I work with VMware ESX since 2004 and never had a single bad experience with VMFS, even in the worst conditions of faulty hardware. VMFS is born as a clustered file system from ground-up, and I can surely affirm that “forcing” NTFS to be a clustered file system is an insane choice, like having the hypervisor layer footprint as big as a Windows 2008 or 2012.
I’d like to know from the Hyper-V experts their experience with upgrading from Hyper-V 2008 to 2012.
In my experience with VMware I did lots of upgrade since ESX 2.x to 3.x to 4.x and to 5.x always without any downtime for running VM (I’d say that the most exciting one was from 2.x to 3.x :-)).
Thanks,
Luca
Your arguments are hardly credible given multiple references to articles published months or years before the release of Windows Sever 2012, as clearly indicated in your links to TechNet blogs.
Seems an amended version of this article with corrections / retractions is in order. A technically flawed rant such as the one published here only call your professionalism and credibility into question.
Pingback: vSphere 5.1 & Hyper-V 2012 memory concepts explained and compared - viktorious.nl - Virtualization & Cloud Management
This is an interesting thread and somewhat amusing. Technically, VMware still more advanced than Hyper-V even with 2012, although, there are significant improvements there.
That said, I don’t know many customers who leverage every vSphere feature that is available to them, but these folks do pay a hefty premium $ for vmware licensing AND microsoft licensing on top of that.
So if you’re the guy who writes the checks, you take a step back and look at the cost vs benefit of vmware vs hyper-v, you simply can not justify the vmware cost. Cost is the reason VMware’s market share has peaked, and it’s for this reason people have started to move workloads to hyper-v. Hyper-V v3 is actually better than “good enough” and certainly has lower costs than vsphere any way you slice it.
Do I care that vSphere has a smaller footprint than Hyper-V? Technical guys might. I don’t. This is not a deal breaker. Do I care to know the details about VMFS vs CSV. Tech guys do. I don’t. What I care about is this – a) Can I run, successfully, my workloads on Hyper-V? b) Can I migrate storage and workloads in the hyper-V cluster as I please c) Can I scale my cluster? d) Is it stable? e) Is it easy to deploy new VMs? e) What are the limits in terms of vCPU and Memory is it good enough for me? f) What are the initial and ongoing costs?
If the answers to all of the above are satisfactory, and in our case, they were, moving to hyper-v is a no brainer.
In my shop we have moved workloads off vSphere and onto Hyper-V and we’re not alone. We did it not because Hyper-V is superior or because we didn’t like VMware. We did because it made financial sense. We did it because we didn’t want to pay premium $ in order to be able to have 2-3 features available which we didn’t even use to begin with and we did it because Hyper-V satisfied most of our technical requirements.
Of course, there are vmware resellers out there who have a vested interest to continue to push vsphere since they make $ off of that, whereas selling microsoft licenses only is not a very profitable business venture.
So while I appreciate the technical merits, at the end of the day people who write the checks, tend to vote with their wallets…
I agree 100% Mike. But because im a tech guy as well as a consultant I care about the same things you talk about plus one more…. its it going to be super easy and stable AFTER I leave. My biggest beef with HyperV is that people advertise it as free… but its not free… there is always a management interface or something that has to be added to the generic windows server install. That part you have to pay for in most cases and from what I hear its not as cheap as people make it sound…. yet no one who has posted a snarky comment here has taken the challenge to post pricing either… I wonder why.
Great blog post Justin!
This is what blogging is all about – giving honest and personal opinion, which is everyone’s right.
After many years supporting and implementing VMware and Hyper-V environments, I cannot help but agree with you regards how solid, straightforward and a pleasure to use VMware is over Hyper-V.
Cheers,
vCosonok
I’d like to know how this clustering and migration issue compares to Red Hat offerings: RHEL (with KVM) + GFS and/or cLVM (cluster suite) or RHEV.
And it would also be nice to find something like this article comparting with Citrix Xen offerings.
I know Linux, be it KVM or Xen, using Red Hat or whatever, has all pieces for a solution but I don’t know easy/hard is to get them working together, and how they perform.
forgot to check “notify me”
as a commercial product i feel as though xenserver has been dead for years
I still see a lot of noise from Citrix abou Xen products on the media and from partners. So, no comments about Xen, KVM or RHEL? All readers here are windows-only? 😉
Sorry didnt mean to ignore that part.
I talked to one of the Ubuntu guys at HP Discover a couple weeks ago, he was one of their openstack guru’s, and i asked him … how many people use openstack to manage other hypervisors (Xen, VMware, etc) and he said almost none… only those who have previously bought the licenses and cant justify not using them yet. He said KVM is by far the majority of what is used these days.
Personally I like Linux a lot, been using it since Debian Potato (2.2 i think or 2.0 cant remember), as for RHEL … i prefer Debian based distros however i use RHEL when i need to for things like vCloud Director and other apps that require it.
No problem, I am curious about products not usually covered by this blog.
Maybe I wasn’t clear. If you know Linux (RHEL in particular, but the same holds for Debian) and KVM, you know KVM can do live migration, and GFS2 provides a clustered file system. So you would be able to create a similar VM cluster setup to the one that was hard to implement and was not very robust using Microsoft Hyper-V and CSV.
“Can do” is not the same as “can do easily” and neither “should run smoothly”. If it runs well, I can tolerate spending some time to do the initial setup, and even spare some time to debug this setup.
I guess RHEV (oVirt, not OpenStack) would make it easy to create the setup. Maybe virt-manager (from RHEL) won’t make it so easy, and I’d have to use many different configuration tools, and maybe even edit some files, to create the intended setup. But once the setup is complete, be it the easy way (oVirt) or the hard way (virt-manager + conga + lvm + whatever) will it work well?
The point is: KVM + GFS2 (or another Linux cluster filesystem, say OCFS or glusterFS) would perform in a robust way, like VmWare VMFS, or would it work no so well, like Microsoft CSV?
I was hoping you (Justin) or some reader had prior experience with the Linux VM and Cluster stack t share and compare with VmWare. Or maybe some reader could talk about GFS2 or OCFS architecure and say it’s closer to VMFS or CSV.
We are down to a single xenserver left. We originally used it to virtualize linux, because it wasnt well supported by Hyper-V in the early days. Since that have since improved alot, we’ve moved most workloads to hyper-v, and will probably turn off the last xenserver in the next 6 months.
Considering the rate of new features and performance of Hyper-V, and beign a Microsoft shop, it doesnt make sense to not go all in on Hyper-V.
Just FYI to the windows folks there, Xen is not KVM.
Older Red Hat (RHEL) releases used Xen, but newer ones use KVM as the Hypervisor.
RHEV is another Red Hat product based on KVM and oVirt. It is closer to the full vSphere suite, while the virtualization features of RHEL would be closer to the “free” VmWare product. So if you didn’t like Citryx XenServer or older RHEL virtualization you may yet be interested in evaluating newer RHEL relases and RHEV.
Of course the same applies to other Linux products based on KVM and oVirt, which by the way are different from OpenStack and CloudStack.
@Mike:
Well, if you don’t require a full feature set, which is typically in small shops, you can use the VMware Essentials (Plus) bundle. In my opinion the Essentials Plus bundle offers a lot of functionality for a very reasonable price. You only have 3 servers/6 sockets, but that’s good enough for quite a number of VMs. You get vCenter to go with it and a number of pretty useful appliances (vSphere Replication, Data Protection) as well.
Yup, I agree completely
I could not agree more. I have been a VMware admin since ESX 1.x…and GSX before that.
We moved our dev environments over to Hyper V 2008 R2 2 year ago. (2 – 8 node clusters). We got trained as much as we could. We worked with Microsoft and our storage vendor to nail down all of the best practices. The result was a very complicated setup, that was not as stable as Vmware. Hotfixes are the issue if you ask me. You have to hunt them down with Hyper V. You don’t with VMware. Every issue we have had has been caused by a missing hot fix. There are some blogs you can follow that break out the hotfixes for Hyper V, storage and failover clustering…but who has the time to do that? We have about 80 hotfixes applied to our 2008 R2 hyper V hosts.
We are now upgrading those dev clusters to server 2012 and SCVMM 2012. What a nightmare it has been. You cant upgrade a MS cluster because you cant have mixed nodes. You must create a new cluster by buying new hardware or breaking your current cluster and removing a node at a time. There is NO WAY to upgrade with zero down time….you must shut down the running VM’s at some point. Our SCVMM 2008 R2 server was running on Windows 2008 R2 SP1. SCVMM 2012 SP1 wont run on windows 2008 R2 SP1…it requires Windows 2012! We could not upgrade our SCVMM 2008 R2 server, we had to build a new one. Serious fail on the part of Microsoft.
I could not imagine running my production services on Hyper V. Yes 2012 finally supports NIC teaming. It takes a few less steps to build a cluster vs 2008 but still quite a bit more than VMware. It has shared storage live migration (who does that with the size of most VM’s these days) SCVMM 2012 SP1 is slow as hell to pickup changes that have been made in Fail Over clustering (because somethings must be done in FOC). SCVMM is better but you still have 3 tools to manage a Hyper V world that is clustered. Often you are in and out of all of them to fix problems. Confusing at best.
Microsoft support costs are staggering! Vmware support comes with the price of the product.
I´m not sure if it was already posted, and didn´t read all the answers here but there is one more thing that saves you money and a lot of trouble when buying vmware on top of the windows datacenter licenses required in both solutions.
You don´t need internal hard drives anymore. Keep in mind, that if you want to install vmware, you can just plug in a 5$ USB thump drive or an SD card. All modern servers come up with at least one internal usb port (or you can even use external) so you don´t need two SAS drives, the SAS controller and maybe a hotspare drive anymore. How long does it take to reinstall an ESXi? Round about 10-15 minutes if you have all the necessary scripts to not implement the network and storage configuration by hand again. How long does it take to install Hyper-V (2012 or 2008R2 doesn´t make a real difference actually)
All these add-ons simply sum up the costs you have to actually implement a hyper-v installation so the term: “It´s included” just doesn´t count at the end. 🙂
@Phoenix – Thanks for the comment. I agree completely!
Also I should note that the customer has had several other incidents and has officially started moving their blades to VMware. 🙂
apart from using outdated links I’ve found numerous to date links of people struggling with another issue with CSV being data protection; we thought we could rely on MS DPM2012 for backing up our HyperV 2012 failover cluster. We were proven wrong; only after applying several hotfixes on the HyperV hosts the situation became somewat stable, but still not 100% reliable..
Just wish I head read this post from Justin earlier..
Pingback: HyperV CSV versus VMWare VMFS | syshead
Still there are so much dependences. I agree with Justin. I am sorry to say this still VMware is more convenient for all of the Technology.
Weather it is HA or DRS or CSV and so many things.
I am working as consultants on Hyper-V using Windows 2012 but still i am not that much convince with the technology.
I hope i am not wrong yes they improve a lots but they need more in to it for end user.
I run 52 machines on a 3-node 2008 R2 cluster backed by an Equallogic, and I’ve had none of these issues. In a way, it’s the iPhone vs. Android debate. Whichever you “grew up with” is the superior solution, and the way the “other guy” does things is stupid. (For instance, VMWare is pretty stupid in the way it handles thin-provisioned VMDK images, and needing multiple versions of the client installed to manage multiple versions of the host.)
Besides, why are you comparing the free version of Hyper-V to the paid version of VMWare? Let’s see you do VMotion, centralized control or even backups on the free version of VMWare.
You have some good points Tim, I agree that having multiple clients installed is silly, and that things such as vmotion and backups using their API’s will have to change at some point since those are as standard as cruise control on a car these days…
We are currently having issues with three nodes in hyper v cluster (2008 r2) since two week and no one can figure out what is wrong, including MS support. After reboot nodes are not coming back to cluster!! On the other hand we have VMware cluster (5.0) and have almost no issues at all, and their support is awesome. My thought about “why should we use hyper v” took me to this page… With all improvements in 2012 r2, still the architecture remains same as 2k8 r2, like cluster manager, csv coordinator node etc., Any unreliable windows service in coordinator node can bring down your VMs on other nodes. This we had experienced many times. From my experience I can say that it is a (very) bad idea of building critical workload on Hyper V. What we really save by using hyper v?
As a Microsoft rep once quoted, “Hyper-V isn’t quite there yet.” In my humble opinion, Hyper-V is a good 3 years behind VMware in the technology as well as functionality. Microsoft needs to get over the fact they couldn’t buy out or take over VMware years ago and focus on what they do, create mediocre operating systems that require many, many, many patches and fixes that lead to additional problems. You gotta love the marketing people. Just like used car salesmen!
I´d say as long as microsoft uses a windows OS as hypervisor platform, they won´t be able to compete with VMware. When they release a pure hypervisor component (compareable to the almost useless core installation) it might get interesting. I just happened to see a VMware environment with 4.0 hosts running for 1300 days. Would really like to see this with a windows host. 🙂
I agree, except on the 1300 day 4.0 environment… to me that just means that someone wasnt doing their job and patching things as they should have… most i have seen was 900 ish days, and it was a huge pain in the ass to upgrade at that point because so many thing were so far out of date.