[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-users] Recommendations for Virtulization Hardware



On 2012-09-27 03:34, ShadesOfGrey wrote:
On 09/24/2012 08:49 AM, Robin Axelsson wrote:
On 2012-09-24 05:45, ShadesOfGrey wrote:
Sorry for the late response, I've had a lot to digest.
<snip>
The lack of current information about Xen (and KVM) online has been frustrating — especially finding the many proof of concept videos that demonstrated possibilities but offered no real specifics. Looking for specifics, I sought information from gaming and enthusiast sites; I figured finding confirmation of VT-d and AMD-Vi support on such sites would be more likely. However, I found that wasn't often the case. I did determine that ASRock motherboards seem to be the most likely to support VT-d, ASUS least likely (unless equipped with an Intel 'sanctioned' VT-d chipset). I had narrowed my choices to two motherboards that appear to offer VT-d support and was intending to contact the manufacturer before purchase. Both choices are a bit pricey and I've been reconsidering whether I should look to other motherboards to reduce costs.

Some motherboards support IOMMU even though it is not found in the user manual or specified on the website. Your best bet is to ask customer support. A guy posted here that he got it working on an Intel motherboard that doesn't even have options for it in the BIOS, so it seems that in some cases it is only up to the CPU. This is not the case with AMD though as I stated before. I have bought a couple of Gigabyte GA990FX-UD7 myself, they are stable and have a good layout. They have support for IOMMU but I haven't tested it thoroughly enough to fully confirm this although I don't believe there would be any problem.

I'm aware of this.  In fact, I only have anecdotal evidence that the Gigabyte G1.Sniper 3 has IOMMU (VT-d) support.  I intend to query the manufacturers, seeking confirmation of IOMMU support, of every motherboard that ends up on my short list.  I may include other Z77 motherboards from MSI and Gigabyte, since there is concrete evidence that the Z77 chipset does support IOMMU.  I did focus, however, on those motherboards I had some inkling could support IOMMU.  Now I'm trying to expand my selection process to include less expensive options.
It surprises me that ASRock and ASUS are so different. ASRock is, or at least used to be a subsidiary of ASUS so there shouldn't be that much difference between them.
Same here.  But I guess things changed after ASRock was spun-off.

<snip>
This is precisely the kind of information I was looking for from the threads I started on Ars Technica. It's just unfortunate that FLR and D3 D0 support aren't often found in the tech specs of must expansion hardware. However, now that I know what to ask, I'll try contacting hardware manufacturers prior to purchasing any expansion hardware. Thank you!

D3 and D0 are power states defined for devices in the ACPI specification and can be used to control the supply voltage (Vcc) to PCI and PCIe devices. You can find more information about it here for example:

http://en.wikipedia.org/wiki/Advanced_Configuration_and_Power_Interface

---------------

Device states

The device states D0-D3 are device-dependent:

  • D0 Fully On is the operating state.
  • D1 and D2 are intermediate power-states whose definition varies by device.
  • D3 Off has the device powered off and unresponsive to its bus.
---------------

So, either it works for a certain type of hardware or it doesn't and I wouldn't expect a vendor to state this "support" in the specifications since it isn't a "feature" in and of itself if you get me. But maybe this will change and maybe FLR support will become more widespread.
I did read that Wikipedia entry after your referenced excerpt.  To my mind, if a manufacturer claims support for the ACPI or PCIe spec and doesn't implement certain portions of those specs(some are optional after all), said manufacturer has an obligation to make that clear to your customers and users.  But then, that's just me.  Anyway, this information gives me what I need to ask the right questions of tech support before purchasing.

I think most motherboards should support this.

I've also had it confirmed by nVidia that they support FLR on all current Quadro cards greater than or equal to Quadro 2000, on Tesla C2050 and higher and M2050 and higher, and on new VGX and GRID cards.

<snip>
From everything I've read, solutions that rely on any form of remote display protocols would be limited to a subset of Direct3D functions. Furthermore, these would vary from one implementation to another, thus making them far less attractive for gaming than VGA passthrough... Well, in my opinion anyway.

VirtualBox's seamless mode is pretty nifty. But it's a Type 2 Hypervisor and relies on paravirtualized drivers that also suffer from the same limitations as remote display protocols. It's great for most things, but gaming is not one of them. And I'm speaking from personal experience. Though I haven't used them myself, the same would seem to hold true of Parallel's and VMWare's 'Workstation' offerings. At least, as far as I've gathered.

FYI, the Type 1 Hypervisors from Parallel's and VMWare* are priced waaayyy outside my budget.

I understand that you want full 3D functionality for Windows gaming but maybe you'll find the subset of 3D functionality for the Linux machine acceptable. I have looked into VirtualGL and with TurboVNC, you might get a pretty decent desktop environment and it seems like most of the features are there already. It appears that the 3D is rendered by hardware/GPU before it is streamed through VNC or Spice. So it seems that you would need another GPU for that. You can find more info on VirtualGL here:

http://www.virtualgl.org/
Upon further investigation, the only option that is available to remotely translate Windows 3D apps is Microsoft's own RemoteFX.  In which case, Microsoft products would be the foundation of my software stack...  Something I'm trying to avoid.
I was just giving you the options I could think of OTOH that would let you share a Windows desktop with a Linux desktop, or rather share the desktop of one VM with the host machine.


Also the line between a type 1 and type 2 hypervisor tend to get a bit blurry. The point with type 1 is that it has access to ring-0 so that it can get access directly to the hardware to be passed through to the guests (I did confuse 'host' and 'guest' in my prior post). It also doesn't need to ask the host OS for permission in the same way as a type 2 hypervisor which is likely to give performance advantages in some cases.

However, even a type 2 hypervisor, although it is run as an application inside the OS can get "type 1" like privileges. By patching into the kernel and/or using special "dummy drivers" for hardware to be shared with VMs you can achieve pretty much the same thing, ergo it is no longer clear whether the hypervisor is a type 1 or type 2.

There is an article about it from the old IBM Mainframe days but I can't seem to find it.

You might be right, the performance of a Type 2 Hypervisor may be sufficient.  Regardless, I'd still have roughly the same hardware requirements as if I were going to use a Type 1 Hypervisor.  I'd still need the same CPU, RAM, GPU, and storage requirements as I've already put forth.  I could omit the potentially necessary additional components for use with a Type 1 Hypervisor.  Things like USB controller, NIC, or sound card.  But those costs would be replaced with the cost of licensing the Type2 Hypervisor.

In which case, I don't really gain or lose anything by experimenting with Type 1 Hypervisors...  At any rate, I could compare and contrast the performance of Type 1 and Type 2 Hypervisors using Paralles' and VMWare's trialware.  It might take a good long while, but it would be an adventure.

*I only found out about VMWare's 'free' vSphere after I'd written this response.

<snip>
Also, it is highly recommended that you use ECC RAM for such applications and it doesn't hurt to dedicate a few gigs of it to the ZFS as RAM is used for cache. The good news is that most motherboards with good chipsets support ECC RAM even though you might not find anything about it in the user manuals.
Again, thanks for the thorough explanation. This gives me a great deal to think about. The more I learn about ZFS, the less appealing it becomes. And by that I mean the confusion over which version of ZFS is in what OS? And just how well maintained the OSes supporting ZFS are? Now I have additional hardware considerations to keep in mind that may (or may not) make the cost of ZFS RAID-Z pool comparable to a hardware RAID5/6 solution anyway. Do you have any suggestions as to which of LSI HBAs I should be considering? I haven't found an HCL for ZFS in my searches.

Out of curiosity — and if you would happen to know — do you think what you suggest about the HBA and SAS drives for ZFS also applies to Btrfs? I'm assuming it would, but I'd appreciate some confirmation.

It's funny how the "I" in RAID never really seems to apply... Especially since it looks more and more like using ZFS or Btrfs will require I commit myself, from the start, to one or the other and a discrete HBA. Transitioning from an integrated SATA controller(s) and mdadm seems rather impractical. If I understand what's involved in doing so correctly. It may turn out that anything other than mdadm is price prohibitive.


I don't think you will have a problem with getting ZFS to run and if that's your only goal then you don't need to be very picky with your choice of hardware. I find ZFS pretty easy and handy to use. I has really great functionality and I don't have many bad things to say about it so far. ZFS is a filesystem (along with a couple of software tools to administrate it) just like EXT4 or NTFS so hardware support depends on the platform it runs on.

But the point with using ZFS is to get maximum protection against data corruption and that's where the selection of hardware gets limited and there are "best practices" set up to achieve that. I have not tested ZFS on any other platform than on OpenSolaris and OpenIndiana but I do know that it is well implemented on that platform and more mature there than on any other (non-solaris) platform. Another advantage with the OSOL/OI platform is that the CIFS functionality is implemented in the kernel space and not in the userland which will give advantages performance wise if you intend to share files with other windows computers. (I don't deny that Samba is pretty good on Linux too. There are some benchmarks on the phoronix website comparing samba with NFS and they are in favor of Samba on those benchmarks...) The second best implementation is found with FreeBSD and it is probably fairly mature but I haven't tested it myself and some people have run into problems with it in the past. The Linux version is probably merely at infancy stage and likely not yet mature enough for regular use. It is probably not as "bad" as btrfs though. There is quite a bit of information about it on the phoronix.com website (and probably also at lwn.net):

http://www.phoronix.com/scan.php?page=news_item&px=MTE4Nzc

My goal wasn't just to experiment with Btrfs, mdamd, or ZFS, if that's what you meant by, "with getting ZFS to run and if that's your only goal".  I actually intend to use it 'in production'.  In fact, the secondary role (Linux desktop being primary) for my proposed Virtualization rig is as file server (incl. httpd).  Windows gaming is a tertiary or quaternary concern for me.  So, finding out that the "best practices" for implementing ZFS include hardware I hadn't anticipated, is a bit off-putting.

The diversity in ZFS implementations is what I meant about being confused.  I'm not as familiar with the underlying platforms that utilize ZFS (other than Linux, and not with ZFS in use).  Without that familiarity, it makes it more difficult to gauge which of those platforms would suit my purposes.  For example, I had read a little bit about the degraded CIFS performance on FreeBSD and Linux due to their reliance on Samba (it residing in user-space being the issue).

BTW, I meant to ask how you came to the conclusion that ECC RAM is supported on desktop motherboards?  It's always been my understanding that you could use ECC RAM on such hardware, but there was no added benefit.
When I looked at the Gigabyte GA990FX motherboard there was no documentation about it anywhere, nor was it stated in the specifications. When I contacted customer support and asked about it, they sent me screen dumps of BIOS showing that it in fact does support ECC providing different ECC scrubbing options. So you can ask support if you are unsure.

The benefit of ECC is that there is a parity bit that checks the integrity of the data that is present in the RAM. We have cosmic radiation and bit flips do occur, also the memory can turn out to be faulty. It has been debated whether we really "need" this on a desktop. The finer litography of the hardware is likely to make it more sensitive to such failures and bit flips than before so it makes more sense to use ECC today than in the past.  If you are unlucky, you computer will crash because of that. Some files may get corrupted in the process. It may not happen very often but it does happen occasionally, hopefully, the bit flips happen in an address space that currently is not in use. If you run a server on the other hand then you are likely to run into freezes and crashes eventually (it may depend on how long uptimes were talking about). Using ECC RAM will prevent those crashes and add extra protection. In fact, Microsoft recommends ECC RAM even on desktop computers.

ZFS does offer protection against data corruption but if you are unlucky, some data corruption may go by undetected when written from RAM. ZFS (or any other filesystem) will write the damaged data to disk and be unable to automatically detect the corruption. This is why using ECC RAM with ZFS is "strongly recommended" in the best practices.

A search there on ZFS will give more articles. The latest official version of ZFS is 28 and is probably implemented in both Linux, and FreeBSD by now. Later versions have been released since Oracle killed the OpenSolaris project and can be found with the commercial closed-source Solaris platform that is supplied by Oracle. Things have happened since Oracle pulled the plug on OSOL project and leading developers behind the ZFS project such as Jeff Bonwick left Sun (after the acquisition by Oracle) and joined up with the Illumos team instead. So you cannot determine the stability of ZFS and ZPOOL merely by looking into the version number unfortunately and I wouldn't expect the FreeBSD implementation to be as stable as the Solaris implementation. It just takes time for the implementation to mature and the bugs to be weeded out and it just happens to have been around for Solaris/OpenSolaris/Illumos for much longer than the other platforms and the Solaris/Illumos version also happens to get first dibs on the features. Among the Illumos people there is an ambition to drop the version numbering altogether and instead talk about available features.

One of the two ZFS implementations on Linux has reached version 28, the other is at version 23 and seems to be abandoned or stagnant (last release was May 2011).  I'm not quite sure what the status of ZFS on FreeBSD is.  From this table on Wikipedia, ZFS is at version 28.  However the table notes that there is no CIFS or iSCSI support, which I'll try to independently confirm.  And, as you say, the fact that ZFS (and Solaris as a whole) has essentially been forked, with the re-consolidation of Solaris as closed-source, just adds to the confusion.  It's possible any given build of Illumos' version 28 of ZFS could have features or bug fixes no present in Oracle's version 33.
I don't know much about FreeBSD but in Solaris/Illumos the CIFS is implemented in the kernel space, in platforms such as Linux and BSD it is provided by the Samba framework in the userspace. If you want iSCSI then you have full support for it in Illumos based operating systems.

Also keep in mind that later version storage pools are not readable on systems with lower versions of zfs/zpool. zpool 28 is implemented in FreeBSD 9.0 and onwards. To my knowledge the ZFSonLinux project also has implemented zpool version 28.

It should be noted that people also have had problems with migrating a ZFS pool from one platform to another even though the version numbers matched.
The recommendation to use SAS hard drives is not so much about the quality of the hard drives themselves as it is about the SAS protocol. The SAS protocol simply handles SCSI transport commands in a better and more reliable manner than do SATA. I believe any decent SAS drive would do. As for HBAs I wrote a list with LSI based hardware a while ago here:

https://www.illumos.org/boards/1/topics/572

the thing is that a lot of OEMs such as IBM, HP, Cisco, Fujitsu-Siemens, Dell, ... supply their branded HBAs with LSI circuitry on them. What hardware to choose depends on what you're looking for. If you want an 8-port controller I would go for Intel SASUC8I or LSI SAS3801E-R. If you want SAS/SATA3 with 6.0 Gb/s then LSI's LSI SAS3801E-R series cards would be a better choice. I don't know what OEMs have come up with in the SATA3 department since I wrote that list but the chips to look for in that case are the LSI MegaRAID 2004/2008/2016e depending on how many ports you want.

If you want to read a further discussion about reliability of different RAID setups I made a post about this in the following thread (last post):

http://communities.intel.com/thread/25945

The cost of the drives is marginally higher than the drives I had budgeted, but including the cost of an SAS HBA is problematic.  I hadn't expected purchasing an HBA from the very start.  I'd hoped to deffer such a purchase for a bit, but you are saying that would be ill-advised.  How likely is it I would encounter problems using SATA drives in the interim?


The problem with "normal" hardware is that the error handling is internal. If a hard drive is starting to have bad sectors, the drive will handle them internally and reallocate the bad sectors. You will not notice anything and the errors will not be reported to the system. At most you will experience sluggish performance, perhaps a freeze. When things go so bad that errors are starting to become visible, then it is too late to do anything about it, the drive is already dead. In the past the drives would return "garbage", in the modern days the drive will reread the sector over and over again while trying to recalibrate the head which gives rise tio the characteristic click-of-death noise.

If you use a "cheap" HBA with say ZFS, the tools for monitoring the hardware such as iostat will be pretty much useless. Things will 'look' just fine with no errors and when errors start to show, they will happen out of the blue with no prior warning. The system drive of a ZFS file server that I run just crashed like that without warning. Prior to that, all I had on the drive were some corrupt blocks of the hard drive image of a virtual machine. It was connected to the Southbridge OnChip SATA controller of the motherboard. I didn't cry about it although there were some system files that I would have liked to keep though. I replaced it with a server grade SATA drive (WD RE4). I have suffered corruption on that drive on one of the virtual hard drive images. Since I don't see much sense in using RAID on the system drive (other people may beg to differ) I enabled the "ditto blocks" feature on that drive and recovered the image file using 'ddrescue'. The drive is doing fine ever since.

On the storage pool of that system I have already replaced two SATA drives on my SAS controller bacause they were going to die. What happened was that the server started to freeze but I couldn't tell what was the problem. Everything looked ok with iostat, was it a driver issue? I couldn't tell. Then eventually I started to see 'media errors' with iostat and when I replaced the drive the pool didn't freeze anymore. Then the same thing happened again and I replaced another hard drive like that. That's the reality of that, if the drives were SAS I would have been able to tell what's wrong at an earlier stage. There was a discussion about this about 6 months ago on the openindiana-discuss mailing list.

Have you seen any price/feature advantage to reseller versions of LSI OEM products?  I would think the offerings from Cisco, Dell, HP, and IBM would come at a premium as opposed to purchasing LSI branded hardware.  Anyway, I thought there might be specific models from particular manufacturers to meet the "best practices" for ZFS.  However, from your suggestions, I gather it really doesn't matter as long as the HBA is Intel or LSI?


That's something you have to look for at your dealers. My experience is that the OEM versions generally are cheaper than LSI branded products. You can compare yourself here:

Intel SASUC8I
http://amzn.to/QjtOR9

LSI 3081E-R
http://amzn.to/USps9x

Both are essentially the same card but the original LSI one is more expensive than the one from Intel.


I wonder, should we perhaps move the discussion of ZFS to private email?  It is a bit off-topic.

I think a better idea is to take this to the openindiana-discuss mailing list and/or the FreeNAS mailing lists/forums as other people with ZFS experience will read your posts. If you post on the openindiana-discuss I will read your post there as well.


_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxx
http://lists.xen.org/xen-users

_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxx
http://lists.xen.org/xen-users

 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.