My last two posts on Microsoft were filled with angst and despair at Microsoft’s announcement that the next gen versions of Server & System Center would be delayed until sometime in 2016. Why, I cried out, why the delay on Server, and what’s to become of my System Center, I wondered?
Well, I was wrong on all that, or perhaps I was only a little bit right.
There was a shakeup, but it wasn’t Nadella who had angrily overturned a gigantic redwood table at System Center HQ, spilling Visio shapes & System Center management packs as he did so, rather it was Mr Windows himself, the Most Distinguished of Distinguished Technical Fellows, Dr. Jeffrey Snover who had shaken things up.
Yes. The Padre of Powershell himself filled in the gaps for me on why System Center & Windows Server were delayed during a TechDays online one day after my last post.
During that talk, he announced that the Windows Server Team has been meshed with the System Center Team and, even better, the Azure team. Hot dog.
[Snover] explained that the System Center team and the Windows Server team are now “a single organization,” with common planning and scheduling. He said that the integration of the two formerly separate organizations isn’t 100 percent, but it’s better than it’s been in the past. The team also takes advantage of joint development efforts with the Microsoft Azure team, he added.
That’s outstanding news in my view.
Microsoft’s private|hybrid|public cloud story is second to none as far as I’m concerned. No one else offers deep integration between cutting edge public cloud systems (Azure) with your on-prem legacy infrastructure stack.
Yet that deep integration (not speaking of AAD Sync & ADFS 3 here) was becoming confused and muddled with overlap between the older tools (System Center) and the newer tools like Desired State Configuration, mixed in with AzurePack, an on-prem/cloud management engine.
It sounds to me like Snover’s going to put together a coherent strategy using all the tools, and I can’t think of a better guy to do the job.
But what of Windows server?
It’s getting Snovered too, but in a way that’s not as clear to me. Again, Redmond mag:
The next Windows Server product will be deeply refactored for cloud scenarios. It will have just the components for that and nothing else, Snover explained. Next, on top of that, Microsoft plans to build a server that will be the same as the Windows Servers that organizations currently use. This server it will have two application profiles. One of the application profiles will target the existing APIs for Windows Server, while the other will target the subsets of the APIs that are cloud optimized, Snover explained. On top of the server, it will be possible to install a client, he added. This redesign is happening to better support automation, he explained.
I watched most of Snover’s talk, took a few days to think about it, and still have no idea what to make of the high-level architecture slide below that flashed on screen briefly:
Some thoughts that ran through my head: is the cloud-optimized server akin to CoreOS, with active/passive boot partitions, something that will finally make Patch Tuesday obsolete? One could hope that with further abstraction, we’ll get something like that in Windows Server vNext.
In some sense, we already have parts of this: if you enable the Hyper-V feature on a bare-metal computer, you emerge, after a few reboots, running a Windows virtual machine atop a Type-1 Hypervisor.
Big deal right? Well, Snover’s slide seems to indicate this will be the default state for the next generation of Windows server, but more than that, it seems to indicate that what we think of as the Type-1 Hyperivisor is getting a bunch of new features, like container support.
We knew Docker support was coming, but at this level, and almost indistinguishable from the hypervisor itself?
That’s potentially all kinds of awesome.
Interestingly, Server Roles & Features look like they’re being recast into a “Client” level that operates above a Windows Server.
Which, if we continue down the rabbit hole, means we have to ask the question: If my AD Domain Controller or my RemoteApp session host farm servers are now clients, what are they running on? It certainly doesn’t seem to be a Windows server anymore, but rather a kind agnostic compute fabric, made up of virtual “Servers” and/or “Containers” operating atop a cloud-optimized server running on bare-metal…an agnostic computing ((Damn straight, had to work that in there)) fabric that stretches across my old on-prem Dells all the way up to the Azure cloud…right?!?
I’m like four levels deep into Jeffrey Snover’s subconscious so I’ll stop, but suffice it to say, the delay of Windows Server & System Center appears to be justified and I can’t wait to start testing it in 2016.
VXLAN, NVGRE & Network Controller, courtesy of Azure: This is something I’ve hoped for in the next version of Windows Server: a more compelling SDN story, something more than Network Function Virtualization & NVGRE encapsulation. If bringing the some of the best -and widely supported- bits of the VMware ecosystem to on-prem Hyper-V & System Center isn’t a virtualization engineer’s wet dream, I don’t know what is.
VMware meet Azure Site Recovery: Coming soon to a datacenter near you, failover your VMware infrastructure via Azure Site Recovery, the same way Hyper-V shops can
In-place/rolling upgrades for Hyper-V Clusters: This feature was announced with the release of Windows Server Technical Preview (of course, I only read about it after I wiped out my lab 2012 R2 cluster) but there’s a lot more detail on it from TechEd via Finn: rebuild physical nodes without evicting them first.You keep the same Cluster Name Object, simply live migrating your VMs off your targeted hosts. Killer.
Single cluster node failure: In the old days, I used to lose sleep over clusres.dll, or clussvc.exe, two important pieces in Microsoft Clustering technology. Sure, your VMs will failover & restart on a new host, but that’s no fun. Ben Armstrong demonstrated how vNext handles node failure by killing the cluster service live during his presentation. Finn says the VMs didn’t failover,but the host was isolated by the other nodes and the cluster simply paused and waited for the node to recovery (up to 4 minutes). Awesome!
Azure Witness: Also for clustering fans who are torn (as I am) between selecting file or disk witness for clusters: you will soon be able to add mighty Azure as a witness to your on-prem cluster. Split brain fears no more!
More enhancements for Storage QoS: Ensure that your tenant doesn’t rob IOPS from everyone else.
The Windows SAN, for real: Yes, we can soon do offsite block-level replication from our on-prem Tiered Storage Spaces servers.
New System Center coming next year: So much to unpack here, but I’ll keep it brief. You may love System Center, you may hate it, but it’s not dead. I’m a fan of the big two: VMM, and ConfigMan. OpsMan I’ve had a love/hate relationship with. Well the news out of TechEd Europe is that System Center is still alive, but more integration with Azure + a substantial new release will debut next summer. So the VMM Technical Preview I’m running in the Daisetta Lab (which installs to C:Program FilesVMM 2012 R2 btw) is not the VMM I was looking for.
Other incredible announcements:
Docker, CoreOS & Azure: Integration of the market-leading container technology with Azure is apparently further along than I believed. A demo was shown that hurts my brain to think about: Azure + Docker + CoreOS, the linux OS that has two OS partitions and is fault-tolerant. Wow
Azure Operational Insights: I’m a fan of the Splunk model (point your firehose of data/logs/events at a server, and let it make sense of it) and it appears Azure Operational Insights is a product that will jump into that space. Screen cap from Finn
This is really exciting stuff.
Looking back on the last few years in Microsoft’s history, one thing stands out: the painful change from the old Server 2008R2 model to the new 2012 model was worth it. All of the things I’ve raved about on this blog in Hyper-V (converged network, storage spaces etc) were just teasers -but also important architectural elements- that made the things we see announced today possible.
The overhaul* of Windows Server is paying huge dividends for Microsoft and for IT pros who can adapt & master it. Exciting times.
* unlike the Windows mobile > Windows Phone transition, which was not worth it
Under the terms of the agreement announced today, the Docker Engine open source runtime for building, running and orchestrating containers will work with the next version of Windows Server. The Docker Engine for Windows Server will be developed as a Docker open source project, with Microsoft participating as an active community member. Docker Engine images for Windows Server will be available in the Docker Hub. The Docker Hub will also be integrated directly into Azure so that it is accessible through the Azure Management Portal and Azure Gallery. Microsoft also will be contributing to Docker’s open orchestration application programming interfaces (APIs).
When I first heard the news, emotion was mixed.
On the one hand, I love it. Virtualization of all flavors -OS, storage, network, and application- is where I want to be, as a blogger, at home in my lab, and professionally.
Yet, as a Windows guy (I dabble, of course), Docker was just a bit out of reach for me, even with my lab, which is 100% Windows.
On the other hand, I also remembered how dreadful it used to be to run Linux applications on Windows. Installing GTK+ Libraries on Windows isn’t fun, and the end-result often isn’t very attractive. In my world, keeping the two separate on the application & OS side/uniting them via Kerberos and/or https/rest has always been my preference.
But that’s old world thinking, ladies and gentlemen.
Because you see, this announcement from Microsoft & Docker Inc sounds deep, rich, functional. Microsoft’s going to contribute some of its Server code to the Docker folks, and the Docker crew will help build Container tech into Windows Server and Azure. I’m hopeful Docker will just be another Role in Server, and that Jeffrey Snover’s powershell cmdlets will hook deep into the Docker stuff.
This probably marks the death of App-V, which I wrote about in comparison to Docker just last month, but that’s fine with me.
Docker on Windows marks a giant step forward for Agnostic Computing…do we dare imagine a future in which our application stacks are portable? Today I’m running an application in a Docker Container on Azure, and tomorrow I move it to AWS?
Microsoft says that’s exactly the vision:
Docker is an open source engine that automates the deployment of any application as a portable, self-sufficient container that can run almost anywhere. This partnership will enable the Docker client to manage multi-container applications using both Linux and Windows containers, regardless of the hosting environment or cloud provider. This level of interoperability is what we at MS Open Tech strive to deliver through contributions to open source projects such as Docker.
Inhale it boys and girls because what you smell is the sweet aroma of VMware VMs being removed from the vSphere collective and placed into System Center & Hyper-V’s warm embrace.
Microsoft has released version three of its V2V and P2V assimilator tool:
Today we are releasing the Microsoft Virtual Machine Converter (MVMC) 3.0, a supported, freely available solution for converting VMware-based virtual machines and virtual disks to Hyper-V-based virtual machines and virtual hard disks (VHDs).
With the latest release, MVMC 3.0 adds the ability to convert a physical computer running Windows Server 2008 or above, or Windows Vista or above to a virtual machine running on a Hyper-V host (P2V).
This new functionality adds to existing features available including:
• Native Windows PowerShell capability that enables scripting and integration into IT automation workflows.
• Conversion and provisioning of Linux-based guest operating systems from VMware hosts to Hyper-V hosts.
• Conversion of offline virtual machines.
• Conversion of virtual machines from VMware vSphere 5.5, VMware vSphere 5.1, and VMware vSphere 4.1 hosts to Hyper-V virtual machines.
This couldn’t have come at a better time for me. At work -which is keeping me so busy I’ve been neglecting these august pages- my new Hyper-V cluster went Production in mid-September and has been running very well indeed.
But building a durable & performance-oriented virtualization platform for a small to medium enterprise is only 1/10th of the battle.
If I were a consultant, I’d have finished my job weeks ago, saying to the customer:
Right. Here you go lads: your cluster is built, your VMM & SCCM are happy, and the various automation bits ‘n bobs that make life in Modern IT Departments not only bearable, but fun, are complete
But I’m an employee, so much more remains to be done. So among many other things, I now transition from building the base of the stack to moving important workloads to it, namely:
Migrating and/or replacing important physical servers to the new stack
Shepherding dozens of important production VMs out of some legacy ESXi 5 & 4 hosts and into Hyper-V & System Center and thence onto greatness
So it’s really great to see Microsoft release a new version of its tool.
So if you work in IT, and even better, if you’re in the virtualization space of IT as I am, you have to know that VMworld is happening this week.
VMworld is just about the biggest vCelebration of vTechnologies there is. Part trade-show, part pilgrimage, part vLollapalooza, VMworld is where all the sexy new vProducts are announced by VMware, makers of ESXi, vSphere, vCenter, and so many other vThings.
It’s an awesome show…think MacWorld at the height of Steve Jobs but with fewer hipsters and way more virtualization engineers. Awesome.
And I’ve never been :sadface:
And 2014’s VMworld was a doozy. You see, the vGiant announced a new 2U, four node vSphere & vSAN cluster-in-a-box hardware device called EVO:RAIL. I’ve been reading all about EVO:RAIL for the last two days and here’s what I think as your loyal Hyper-V blogger:
What’s in a name? Right off the bat, I was struck by the name for this appliance. EVO:RAIL…say what? What’s VMware trying to get across here? Am I to associate EVO with the fast Mitsubishi Lancers of my youth, or is this EVO in the more Manga/Anime sense of the word? Taken together, EVO:RAIL also calls to mind sci-fi, does it not? You could picture Lt. Cmdr Data talking about an EVO:RAIL to Cmdr Riker, as in “The Romulan bird of prey is outfitted with four EVO:RAIL phase cannons, against which the Enterprise’s shields stand no chance.” Speaking of guns: I also thought of the US Navy’s Railguns; long range kinetic weapons designed to destroy the Nutanix/Simplivity the enemy.
If you’re selling an appliance, do you need vExperts? One thing that struck me about VMware’s introduction of EVO:RAIL was their emphasis on how simple it is to rack, stack, install, deploy and virtualize. They claim the “hyper-converged” 2U box can be up and running in about 15 minutes; a full rack of these babies could be computing for you in less than 2 hours. They’ve built a sexy HTML 5 GUI to manage the thing, no vSphere console or PowerCLI in sight. It’s all pre-baked, pre-configured, and pre-built for you, the small-to-medium enterprise. It’s so simple a help desk guy could set it up. So with all that said, do I still need to hire vExperts and VCDX pros to build out my virtualization infrastructure? It would appear not. Is that the message VMware is trying to convey here?
One SKU for the Win: I can’t be the only one that thinks buying the VMware stack is a complicated & time-consuming affair. Chris Wahl points out that EVO:RAIL is one SKU, one invoice, one price to pay, and VMware’s product page confirms that, saying you can buy a Dell EVO:RAIL or a Fujitsu EVO:RAIL, but whatever you buy, it’ll be one SKU. This is really nice. But why? VMware is famous for licensing its best-in-class features…why mess with something that’s worked so well for them?
One could argue that EVO:RAIL is a reaction to simplified pricing structures on rival systems…let’s be honest with ourselves. What’s more complicated: buying a full vSphere and/or vHorizon suite for a new four node cluster, or purchasing the equivalent amount of computing units in Azure/AWS/Google Compute? What model is faster to deploy, from sales call to purchasing to receiving to service? What model probably requires consulting help?
Don’t get me wrong, I think it’s great. I like simple menus, and whereas buying VMware stuff before was like choosing from a complicated, multi-page, multi-entree menu, now it’s like buying burgers at In ‘n Out. That’s very cool, but it means something has changed in vLand.
I love the density: As someone who’s putting the finishing touches on my own new virtualization infrastructure, I love the density in EVO:RAIL. 2 Rack Units with E5-26xx class Xeons packing 6 cores each means you can pack about 48 cores into 2U! Not bad, not bad at all. The product page also says you can have up to 16TB of stroage in those same 2U (courtesy of VSAN) and while you still need a ToR switch to jack into, each node has 2x10GbE SFP+ or Copper. Which is excellent. RAM is the only thing that’s a bit constrained; each node in an EVO:RAIL can only hold 192GB of RAM, a total of 768GB per EVO:RAIL.In comparison, my beloved 2U pizza boxes offer more density in some places, but less overall, given than 1 Pizza Box = one node. In the Supermicros I’m racking up later this week, I can match the core count (4×12 Core E5-46xx), improve upon the RAM (up to 1TB per node) and easily surpass the 16TB of storage. That’s all in 2U and all for about $15-18k.Where the EVO:RAIL appears to really shine is in VM/VDI density. VMware claims a single EVO:RAIL is built to support 100 General Purpose VMs or to support up to 250 VDI sessions, which is f*(*U#$ outstanding.
I wonder if I can run Hyper-V on that: Of course I thought that. Because that would really kick ass if I could.
Overall, a mighty impressive showing from VMware this week. Like my VMware colleagues, I pine for an EVO:RAIL in my lab.
I think EVO:RAIL points to something bigger though…This product marks a shift in VMware’s thinking, a strategic reaction to the changes in the marketplace. This is not just a play against Nutranix and other hyper-converged vendors, but against the simplicity and non-specialist nature of cloud Infrastructure as a Service. This is a play against complexity in other words…this is VMware telling the marketplace that you can have best-in-class virtualization without worst-in-class licensing pain and without hiring vExperts to help you deploy it.
A few brief updates & random thoughts from the last few days on all the stuff I’ve been working on.
Refreshing the Core at work: Summer’s ending, but at work, a new season is advancing, one rack unit at a time. I am gradually racking up & configuring new compute, storage, and network as it arrives; It Is Not About the Hardware™, but since you were wondering: 64 Ivy Bridge cores and about 512GB RAM, 30TB of storage, and Nexus 3k switching.
Ahh, the Nexus line. Never had the privilege to work on such fine switching infrastructure. Long time admirer, first-time NX-OS user. I have a pair of them plus a Layer 3 license so the long-term thinking involves not just connecting my compute to my storage, but connecting this dense stack northbound & out via OSPF or static routes over a fault-tolerant HSRP or VRRP config.
To do that, I need to get familiar with some Nexus-flavored acronyms that aren’t familiar to me: virtual port channels (VPC), Control Plane policy (COPP), VRF, and oh-so-many-more. I’ll also be attempting to answer the question once and for all: what spanning tree mode does one use to connect a Nexus switch to a virtualization host running Hyper-V’s converged switching architecture? I’ve used portfast in the lab on my Catalyst, but the lab switch is five years old, whereas this Nexus is brand new. And portfast never struck me as the right answer, just the easy one.
To answer those questions and more, I have TAC and this excellent tome provided gratis by the awesome VAR who sold us much of the equipment.
Into the vCPU Blender goes Lync: Last Friday, I got a call from my former boss & friend who now heads up a fast-growing IT department on the coast. He’s been busy refreshing & rationalizing much of his infrastructure as well, but as is typical for him, he wants more. He wants total IT transformation, so as he’s built out his infrastructure, he laid the groundwork to go 100% Microsoft Lync 2013 for voice.
Yeah baby. Lync 2013 as your PBX, delivering dial tone to your endpoints, whether they are Bluetooth-connected PC headsets, desk phones, or apps on a mobile.
Forget software-defined networking. This is software-defined voice & video, with no special server hardware, cloud services, or any other the other typical expensive nonsense you’d see in a VoIP implementation.
If Lync 2013 as PBX is not on your IT Bucket List, it should be. It was something my former boss & I never managed to accomplish at our previous employer on Hyper-V.
Now he was doing it alone. On a fast VMware/Nexus/NetApp stack with distributed vSwitches. And he wanted to run something by me.
So you can imagine how pleased I was to have a chat with him about it.
He was facing one problem which threatened his Go Live date: Mean Opinion Score, or MOS, a simple 0-5 score Lync provides to its administrators that summarizes call quality. MOS is a subset of a hugely detailed Media Quality Summary Report, detailed here at TechNet.
My friend was scoring a .6 on his MOS. He wanted it to be at 4 or above prior to go-live.
So at first we suspected QoS tags were being stripped somewhere between his endpoint device and the Lync Mediation VM. Sure enough, Wireshark proved that out; a Distributed vSwitch (or was it a Nexus?) wasn’t respecting the tag, resulting in a sort of half-duplex QoS if you will.
He fixed that, ran the test again, and still: .6. Yikes! Two days to go live. He called again.
That’s when I remembered the last time we tried to tackle this together. You see, the Lync Mediation Server is sort of the real PBX component in Lync Enterprise Voice architecture. It handles signalling to your endpoints, interfaces with the PSTN or a SIP trunk, and is the one server workload that, even in 2014, I’d hesitate making virtual.
My boss had three of them. All VMs on three different VMware hosts across two sites.
I dug up a Microsoft whitepaper on virtualizing Lync, something we didn’t have the last time we tried this. While Redmond says Lync Enterprise Voice on top of VMs can work, it’s damned expensive from a virtualization host perspective. MS advises:
You should disable hyperthreading on all hosts.
Do not use processor oversubscription; maintain a 1:1 ratio of virtual CPU to physical CPU.
Make sure your host servers support nested page tables (NPT) and extended page tables (EPT).
Disable non-uniform memory access (NUMA) spanning on the hypervisor, as this can reduce guest performance.
Talk about Harshing your vBuzz. Essentially, building Lync out virtually with Enterprise Voice forces you to go sparse on your hosts, which is akin to buying physical servers for Lync. If you don’t, into the vCPU blender goes Lync, and out comes poor voice quality, angry users, bitterness, regret and self-punishment.
Anyway, he did as advised, put some additional vCPU & memory reservations in place on his hosts, and yesterday, whilst I was toiling in the Hot Lane, he called me from Lync via his mobile.
He’s a married man just like me, but I must say his voice sounded damn sexy as it was sliced up into packets, sent over the wire, and converted back to analog on my mobile’s speaker. A virtual chest bump over the phone was next, then we said goodbye.
Another Go Live Victory (by proxy). Sweet.
Azure Outage: Yesterday’s bruising hours-long global Azure outage affected Virtual Machines, storage blobs, web services, database services and HD Insight, Microsoft’s service for big data crunching. As it unfolded, I navel-gazed, when I felt like helping. There was literally nothing I could do. Had I some crucial IaaS or PaaS in the Azure stack, I’d be shit out of luck, just like the rest. I felt quite helpless; refreshing Mary Jo’s page and the Azure dashboard didn’t help. I wondered what the problem was; it’s been a difficult week for Microsofties whether on-prem or in Azure. Had to be related to the update cycle, I thought.
On the plus side, Azure Active Directory services never went down, nor did several other services. Office 365 stayed up as well, though it is built atop separate-but-related infrastructure in my understanding.
Lastly, I pondered two thoughts: if you’re thinking of reducing your OpEx by replacing your DR strategy with an Azure Site Recovery strategy, does this change your mind? And if you’re building out Azure as your primary IaaS or PaaS, do you just accept such outages or do you plan a failback strategy?
Labworks : Towards a 100% Windows-defined Daisetta Lab: What’s next for the Daisetta Lab? Well, I have me an AMD Duron CPU, a suitable motherboard, a 1U enclosure with PSU, and three Keepin’ it RealTek NICs. Oh, I also have a case of the envies, envies for the VMware crowd and their VXLAN and NSX and of course VMworld next week. So I’m thinking of building a Network Virtualization Gateway appliance. For those keeping score at home, that would mean from Storage to Compute to Network Edge, I’d have a 100% Windows lab environment, infused with NVGRE which has more use cases than just multi-tenancy as I had thought.
This is a really lame but (IMHO) effective drawing of what I think of as a modern small/medium business enterprise ‘stack’:
As you can see, just about every element of a modern IT is portrayed.
Down at the base of the pyramid, you got your storage. IOPS, RAID, rotational & ssd, snapshots, dedupes, inline compression, site to site storage replication, clones and oh me oh my…all the things we really really love are right here. It’s the Luntastic layer and always will be.
Above that, your compute & Memory. The denser the better, 2U Pizza Boxes don’t grow on trees and the business isn’t going to shell out more $$$ if you get it wrong.
Above that, we have what my networking friends would call the “Underlay network.” Right. Some cat 6, twinax, fiber, whatever. This is where we push some packets, whether to our storage from our compute, northbound out to the world, southbound & down the stack, or east/west across it. Leafs, spines, encapsulation, control & data planes, it’s all here.
And going higher -still in Infrastructure Land mind you- we have the virtualization layer. Yeah baby. This is what it’s all about, this is the layer that saved my career in IT and made things interesting again. This layer is designed to abstract all that is beneath it with two goals in mind: cost savings via efficiency gains & ease of provisioning/use.
And boy,has this layer changed the game, hasn’t it?
So if you’re a virtualization engineer like I am, maybe this is all you care about. I wouldn’t blame you. The infrastructure layer is, after all, the best part of the stack, the only part of the stack that can claim to be #Glorious.
But in my career, I always get roped in (willingly or not) into the upper layers of the stack. And so that is where I shall take you, if you let me.
Next up, the Platform layer. This is the layer where that special DBA in your life likes to live. He optimizes his query plans atop your Infrastructure layer, and though he is old-school in the ways of storage, he’s learned to trust you and your fancy QoS .vhdxs, or your incredibly awesome DRS fault-tolerant vCPUs.
Or maybe you don’t have a DBA in your Valentine’s card rotation. Maybe this is the layer at which the devs in your life, whether they are running Eclipse or Visual Studio, make your life hell. They’re always asking for more x (x= memory, storage, compute, IP), and though they’re highly-technical folks, their eyes kind of glaze over when you bring up NVGRE or VXLAN or Converged/Distributed Switching or whatever tech you heart at the layer below.
Then again, maybe you work in this layer. Maybe you’re responsible for building & maintaining session virtualization tech like RDS or XenApp, or maybe you maintain file shares, web farms, or something else.
Point is, the people at this layer are platform builders. To borrow from the automotive industry, platform guys build the car that travels on the road infrastructure guys build. It does no good for either of us if the road is bumpy or the car isn’t reliable, does it? The user doesn’t distinguish between ‘road’ and ‘car’, do they? They just blame IT.
Next up: software & service layer. Our users exist here, and so do we. Maybe for you this layer is about supporting & deploying Android & iPhone handsets and thinking about MDM. Or maybe you spend your day supporting old-school fat client applications, or pushing them out.
And finally, now we arrive to the top of the pyramid. User-space. The business.
This is where (and the metaphor really fits, doesn’t it?) the rubber meets the road ladies and gentlemen. It’s where the business user drives the car (platform) on the road (infrastructure). This is where we sink or swim, where wins are tallied and heros made, or careers are shattered and the cycle of failure>begets>blame>begets>fear>begets failure begins in earnest.
That’s the stack. And if you’re in IT, you’re in some part of that stack, whether you know it or not.
But the stack is changing. I made a silly graphic for that too. Maybe tomorrow.
My idea was to combine all types of disks -rotational 3.5″ & 2.5″ drives, SSDs, mSATAs, hell, I considered USB- into one tight, well-built storage box for my lab and home data needs. A sort of Storage Ark, if you will; all media types were welcome, but only if they came in twos (for mirroring & Parity sake, of course) and only if they rotated at exactly 7200 RPM and/or leveled their wears evenly across the silica.
And onto this unholy motley crue of hard disks I slapped a software architecture that promised to abstract all the typical storage driver, interface, and controller nonsense away, far, far away in fact, to a land where the storage can be mixed, the controllers diverse, and by virtue of the software-definition bits, network & hypervisor agnostic. In short, I wanted to build an agnostic #StorageGlory box in the Daisetta Lab.
Right. So what did I use to achieve this? ZFS and Zpools?
Hell no, that’s so January.
VSAN? Ha! I’m no Chris Wahl.
I used Windows, naturally.
That’s right. Windows. Server 2012 R2 to be specific, running Core + Infrastructure GUI with 8GB of RAM, and some 17TB of raw disk space available to it. And a little technique developed by the ace Microsoft server team called Tiered Storage Spaces.
Was a #StorageGlory Achievement Unlocked, or was it a dud?
Here’s my review after 30 days on my Windows SAN: san.daisettalabs.net.
It doesn’t make you pick a side in either storage or storage-networking: Do you like abstracted pools of storage, managed entirely by software? Put another way, do you hate your RAID controller and crush on your old-school NetApp filer, which seemingly could do everything but object storage?
When I say block, do you instinctively say file? Or vice-versa?
Well then my friend, have I got a storage system for your lab (and maybe production!) environment: Windows Storage Spaces (now with Tiering!) offers just about everything guys like you or me need in a storage system for lab & home media environments. I love it not just because it’s Microsoft, but also because it doesn’t make me choose between storage & storage-networking paradigms. It’s perhaps the ultimate agnostic storage technology, and I say that as someone who thinks about agnosticism and storage.
You know what I’m talking about. Maybe today, you’ll need some block storage for this VM or that particular job. Maybe you’re in a *nix state of mind and want to fiddle with NFS. Or perhaps you’re feeling bold & courageous and decide to try out VMware again, building some datastores on both iSCSI LUNs and NFS shares. Then again, maybe you want to see what SMB 3.0 3.0 is all about, the MS fanboys sure seem to be talking it up.
The point is this: I don’t care what your storage fancy is, but for lab-work (which makes for excellence in work-work) you need a storage platform that’s flexible and supportive of as many technologies as possible and is, hopefully, software-defined.
And that storage system is -hard to believe I’ll grant you- Windows Server 2012 R2.
I love storage and I can’t think of one other storage system -save for maybe NetApp- that let’s me do crazy things like store .vmdks inside of .vhdxs (oh the vIrony!), use SMB 3 multichannel over the same NICs I’m using for iSCSI traffic, create snapshots & clones just like big filers all while giving me the performance-multiplier benefits of SSDs and caching and a reasonable level of resiliency.
File this one under WackWackStorageGloryAchievedWindows boys and girls.
I can do it all with Storage Spaces in 2012 R2.
As I was thinking about how to write about Storage Spaces, I decided to make a chart, if only to help me keep it straight. It’s rough but maybe you’ll find it useful as you think about storage abstraction/virtualization tech:
It’s easy to build and supports your disks & controllers: This is a Microsoft product. Which means it’s easy to deploy & build for your average server guy. Mine’s running on a very skinny, re-re-purposed SanDisk Ready Cache SSD. With Windows 2012 R2 server running the Infrastructure Management GUI (no explorer.exe, just Server Manager + your favorite snap-ins), it’s using about 6GB of space on the boot drive.
And drivers for the Intel C226 SATA controller, the LSI 9218si SAS card, and the extra ASMedia 1061 controller were all installed automagically by Windows during the build.
The only other system that came close to being this easy to install -as a server product- was Oracle Solaris 11.2 Beta. It found, installed drivers for, and exposed all controllers & disks, so I was well on my way to going the ZFS route again, but figured I’d give Windows a chance this time around.
Nexenta 4, in contrast, never loaded past the Install Community Edition screen.
It’s improved a lot over 2012: Storage Spaces almost two years ago now, and I remember playing with it at work a bit. I found it to be a mind-f*** as it was a radically different approach to storage within the Windows server context.
I also found it to be slow, dreadfully slow even, and not very survivable. Though it did accept any disk disk I gave it, it didn’t exactly like it when I removed a USB drive during an extended write test. And it didn’t take the disk back at the conclusion of the test either.
Like everything else in Microsoft’s current generation, Storage Spaces in 2012 R2 is much better, more configurable, easier to monitor, and more tolerant of disk failures.
It also has something for the IOPS speedfreak inside all of us.
Tiered Storage Spaces & Adjustable write cache: Coming from ZFS & the Adaptive Replacement Cache, the ZFS Intent Log, the SLOG, and L2ARC, I was kind of hooked on the idea of using massive amounts of my ECC RAM to function as a sort of poor-mans NVRAM.
Windows can’t do that, but with Tiered Storage Spaces, you can at least drop a few SSDs in your array (in my case three x 256GB 840 EVO & one 128GB Samsung 830), mix them into your disk pool, and voila! Fast read-cache, with a Microsoft-flavored MRU/LFU algorithm of some type keeping your hottest data on the fastest disks and your old data on the cheep ‘n deep rotationals.
Which i naturally did while building this guy out. I mean, who wouldn’t want more write-cache?
But there’s a huge gotcha buried in the Technet and blogposts I found about this. I wanted to pool all my disks together into as large of a single virtual disk as possible, then pack iSCSI-connected .vhdxs, SMB 3 shares, and more inside that single, durable & tiered virtual disk.What I didn’t want was several virtual disks (it helped me to think of virtual disks as a sort of Aggregate) with SMB 3 shares and vhdx files stored haphazardly between them.
Which is what you get when you adjust the write-cache size. Recall that I have a capacity of about 17TB raw among all my disks. Building a storage pool, then a virtual disk with a 10GB write cache gave me a tiered virtual disk with a maximum size of about 965GB. More on that below.
It can be wicked fast, but so is RAID 0: Check out my standard SQLIO benchmark routine, which I run against all storage technologies that come my way. The 1.5 hour test is by no means comprehensive -and I’m not saying the IOPS counter is accurate at all (showing max values across all tests by the way)- but I like this test because it lets me kick the tires on my array, take her out for a spin, and see how she handles.
And with a “Simple” layout (no redundancy, probably equivalent to RAID 0), she handles pretty damn well, but even I’m not crazy enough to run tiered storage spaces in a simple layout config:
What’s odd is how poorly the array performed with 10GB of “Write Cache.” Not sure what happened here, but as you can see, latency spiked higher during the 10GB write cache write phase of the test than just about every other test segment.
Something to do with parity no doubt.
For my lab & home storage needs, I settled on a Mirror 2-way parity setup that gives me moderate performance with durability in mind, though not much as you’ll see below.
Making the most of my lab/home network and my NICs: Recall that I have six GbE NICs on this box. Two are built into the Supermicro board itself (Intel), and the other four come by way of a quad-port Intel I350-T4 server NIC.
Anytime you’re planning to do a Microsoft cluster in the 1GbE world, you need lots of NICs. It’s a bit of a crutch in some respects, especially in iSCSI. Typically you VLAN off each iSCSI NIC for your Hyper-V hosts and those NICs do one thing and one thing only: iSCSI, or Live Migration, or CSV etc. Feels wasteful.
But on my new storage box at home, I can use them for double-duty: iSCSI (or LM/CSV) as well as SMB 3. Yes!
Usually I turn off Client for Microsoft Networks (the SMB file sharing toggle in NIC properties) on each dedicated NIC (or vEthernet), but since I want my file cake & my block cake at the same time, I decided to turn SMB on on all iSCSI vEthernet adapters (from the physical & virtual hosts) and leave SMB on the iSCSI NICs on san.daisettalabs.net as well.
The end result? This:
[table caption=”Storage Networking-All of the Above Approach” width=”500″ colwidth=”20|100|50″ colalign=”left|left|center|left|right”]
1,MGMT,100,192.168.100.15,MGMT & SMB3
2,CLNT,102,192.168.102.15,Home net & SMB3
3,iSCSI-10,10,172.16.10.x,iSCSI & SMB3
4,iSCSI-11,10,172.16.11.x,iSCSI & SMB3
5,iSCSI-12,10,172.16.12.x,iSCSI & SMB3
That’s five, count ’em five NICs (or discrete channels, more specifically) I can use to fully soak in the goodness that is SMB 3 multichannel, with the cost of only a slightly unsettling epistemological question about whether iSCSI NICs are truly iSCSI if they’re doing file storage protocols.
Now SMB 3 is so transparent (on by default) you almost forget that you can configure it, but there’s quite a few ways to adjust file share performance. Aidan Finn argues for constraining SMB 3 to certain NICs, while Jose Barreto details how multichannel works on standalone physical NICs, a pair in a team, and multiple teams of NICs.
I haven’t decided which model to follow (though on san.daisettalabs.net, I’m not going to change anything or use Converged switching…it’s just storage), but SMB 3 is really exciting and it’s great that with Storage Spaces, you can have high performance file & block storage. I’ve hit 420MB/sec on synchronous file copies from san to host and back again. Outstanding!
I Finally got iSNS to work and it’s…meh: One nice thing about san.daisettalabs.net is that that’s all you need to know…the FQDN is now the resident iSCSI Name Server, meaning it’s all I need to set on an MS iSCSI Initiator. It’s a nice feature to have, but probably wasn’t worth the 30 minutes I spent getting it to work (hint: run set-wmiinstance before you run iSNS cmdlets in powershell!) as iSNS isn’t so great when you have…
SMI-S, which is awesome for Virtual Machine Manager fans: SMI-S, you’re thinking, what the the hell is that? Well, it’s a standardized framework for communicating block storage information between your storage array and whatever interface you use to manage & deploy resources on your array. Developed by no less an august body than the Storage Networking Industry Association (SNIA), it’s one of those “standards” that seem like a good idea, but you can’t find it much in the wild as it were. I’ve used SMI-S against a NetApp Filer (in the Classic DoT days, not sure if it works against cDoT) but your Nimbles, your Pures, and other new players in the market get the same funny look on their face when you ask them if they support SMI-S.
“Is that a vCenter thing?” they ask.
Microsoft, to its credit, does. Right on Windows Server. It’s a simple feature you install and two or three powershell commands later, you can point Virtual Machine Manager at it and voila! Provision, delete, resize, and classify iSCSI LUNS on your Windows SAN, just like the big boys do (probably) in Azure, only here, we’re totally enjoying the use of our corpulent.vhdx drives, whereas in Azure, for some reason, they’re still stuck on .vhds like rookies. Haha!
It’s a very stable storage platform for Microsoft Clustering: I’ve built a lot of Microsoft Hyper-V clusters. A lot. More than half a dozen in production, and probably three times that in dev or lab environments, so it’s like second nature to me. Stable storage & networking are not just important factors in Microsoft clusters, they are the only factors.
So how is it building out a Hyper-V cluster atop a Windows SAN? It’s the same, and different at the same time, but, unlike so many other cluster builds, I passed the validation test on the first attempt with green check marks everywhere. And weeks have gone by without a single error in the Failover Clustering snap-in; it’s great.
It’s expensive and seemingly not as redundant as other storage tech: When you build your storage pool out of offlined disks, your first choice is going to involve (just like other storage abstraction platforms) disk redundancy. Microsoft makes it simple, but doesn’t really tell you the cost of that redundancy until later in the process.
Recall that I have 17TB of raw storage on san.daisettalabs.net, organized as follows:
Disk Type, Quantity, Size, Format, Speed, Function
WD Red 2.5″ with NASWARE, 6, 1TB, 4KB AF, SATA 3 5400RPM, Cheep ‘n deep
Samsung 840 EVO SSD, 3, 256GB, 512byte, 250MB/read, Tiers not fears
Samsung 830 SSD, 1, 128GB, 512byte, 250MB/read, Tiers not fears
HGST 3.5″ Momentus, 6, 2TB, 512byte, 105MB/r/w, Cheep ‘n deep
Now, according to my trusty IOPS Excel calculator, if I were to use traditional RAID 5 or RAID 6 on that set of spinners, I’d get about 16.5TB usable in the former, 15TB usable in the latter (assuming RAID penalty of 5 & 6, respectively)
For much of the last year, I’ve been using ZFS & RAIDZ2 on the set of six WD Red 2.5″ drives. Those have a raw capacity of 6TB. In RAIDZ2 (roughly analogous to RAID 6), I recall getting about 4.2TB usable.
All in all, traditional RAID & ZFS’ RAIDZ cost me between 12% and 35% of my capacity respectively.
So how much does Windows Storage Spaces resiliency model (Mirrored, 2-way parity) cost me? A lot. We’re in RAID-DP territory here people:
Ack! With 17TB of raw storage, I get about 5.7TB usable, a cost of about 66%!
And for that, what kind of resieliency do I get?
I sure as hell can’t pull two disks simultaneously, as I did live during prod in my ZFS box. I can suffer the loss of only a single disk. And even then, other Windows bloggers point to some pain as the array tries to adjust.
Now, I’m not the brightest on RAID & parity and such, so perhaps there’s a more resilient, less costly way to use Storage Spaces with Tiering, but wow…this strikes me as a lot of wasted disk.
Not as easy to de-abstract the storage: When a disk array is under load, one of my favorite things to do is watch how the IO hits the physical elements in the array. Modern disk arrays make what your disks are doing abstract, almost invisible, but to truly understand how these things work, sometimes you just want the modern equivalent of lun stats.
In ZFS, I loved just letting gstat run, which showed me the load my IO was placing on the ARC, the L2ARC and finally, the disks. Awesome stuff:
As best as I can tell, there’s no live powershell equivalent to gstat for Storage Spaces. There are teases though; you can query your disks, get their SMART vitals, and more, but peeling away the onion layers and actually watching how Windows handles your IO would make Storage Spaces the total package.
So that’s about it: this is the best storage box I’ve built in the Daisetta Lab. No regrets going with Windows. The platform is mature, stable, offers very good performance, and decent resiliency, if at a high disk cost.
I’m so impressed I’ve checked my Windows SAN skepticism at the door and would run this in a production environment at a small/medium business (clustered, in the Scaled Out File Server role). Cost-wise, it’s a bargain. Check out this array: it’s the same exact Hardware a certain upstart Storage vendor I like (that rhymes with Gymbal Porridge) sells, but for a lot less!
So you may recall that back in March, yours truly, Parent Partition, was invited as a delegate to a Tech Field Day event, specifically Virtualization Field Day #3, put on by the excellent team at Gestalt IT especially for the guys guys who like V.
And you may recall further that as I diligently blogged the news and views to you, that by day 3, I was getting tired and grumpy. Wear leveling algorithms intended to prevent failure could no longer cope with all this random tech field day IO, hot spots were beginning to show in the parent partition and the resource exhaustion section of the Windows event viewer, well, she was blinking red.
And so, into this pity-party I was throwing for myself walked a Russian named Yuri, a Dr. named Schmuel and a product called a “VMTurbo” as well as a Macbook that like all Mac products, wouldn’t play nice with the projector.
You can and should read all about what happened next because 1) VMTurbo is an interesting product and I worked hard on the piece, and 2) it’s one of the most popular posts on my little blog.
Now the great thing about VMTurbo OpsMan & Yuri & Dr. Schmuel’s presentation wasn’t just that it played into my fevered fantasies of being a virtualization economics czar (though it did), or that it promised to bridge the divide via reporting between Infrastructure guys like me and the CFO & corner office finance people (though it can), or that it had lots of cool graphs, sliders, knobs and other GUI candy (though it does).
No, the great thing about VMTurbo OpsMan & Yuri & Dr. Schmuel’s presentation was that they said it would work with that other great Type 1 Hypervisor, a Type-1 Hypervisor I’m rather fond of: Microsoft’s Hyper-V.
And so in the last four or five weeks of my employment with Previous Employer (PE), I had the opportunity to test these claims, not in a lab environment, but against the stack I had built, cared for, upgraded, and worried about for four years.
That’s right baby. I put VMTurbo’s economics engine up against my six node Hyper-V cluster in PE’s primary datacenter, a rationalized but aging cluster with two iSCSI storage arrays, a 6509E, and 70+ virtual machines.
Who’s the better engineer? Me, or the Boston appliance designed by a Russian named Yuri and a Dr. named Schmuel?
Here’s what I found.
Thinking economically isn’t just part of the pitch: VMTurbo’s sales reps, sales engineers and product managers, several of whom I spoke with during the implementation, really believe this stuff. Just about everyone I worked with stood up to my barrage of excited-but-serious questioning and could speak literately to VMTurbo’s producer/consumer model, this resource-buys-from-that-resource idea, the virtualized datacenter as a market analogy. The company even sends out Adam Smith-themed emails (Famous economist…wrote the Wealth of Nations if you’re not aware). If your infrastructure and budget are similar to what mine were at PE, if you stress over managing virtualization infrastructure, if you fold paper again and again like I did, VMTurbo gets you.
Installation of the appliance was easy: Install process was simple: download a zipped .vhd (not .vhdx), either deploy it via VMM template or put the VHD into a CSV and import it, connnect it to your VM network, and start it up. The appliance was hassle-free as a VM; it’s running Suse Linux, and quite a bit of java code from what I could tell, but for you, it’s packaged up into a nice http:// site, and all you have to do is pop in the 30 day license XML key.
It was insightful, peering into the stack from top to nearly the bottom and delivering solid APM: After I got the product working, I immediately made the VMturbo guys help me designate a total of about 10 virtual machines, two executables, the SQL instances supporting those .exes and more resources as Mission Critical. The applications & the terminal services VMs they run on are pounded 24 hours a day, six days a week by 200-300 users. Telling VMTurbo to adjust its recommendations in light of this application infrastructure wasn’t simple, but it wasn’t very difficult either. That I finally got something to view the stack in this way put a bounce in my step and a feather in my cap in the closing days of my time with PE. With VMTurbo, my former colleagues on the help desk could answer “Why is it slow?!?!” and I think that’s great.
Like mom, it points out flaws, records your mistakes and even puts a $$ on them, which was embarrassing yet illuminating: I was measured by this appliance and found wanting. VMTurbo, after watching the stack for a good two weeks, surprisingly told me I had overprovisioned -by two- virtual CPUs on a secondary SQL server. It recommended I turn off that SQL box (yes, yes, we in Hyper-V land can’t hot-unplug vCPU yet, Save it VMware fans!) and subtract two virtual CPUs. It even (and I didn’t have time to figure out how it calculated this) said my over-provisioning cost about $1200. Yikes.
It’s agent-less: And the Windows guys reading this just breathed a sigh of relief. But hold your golf clap…there’s color around this from a Hyper-V perspective I’ll get into below. For now, know this: VMTurbo knocked my socks off with its superb grasp & use of WMI. I love Windows Management Instrumentation, but VMTurbo takes WMI to a level I hadn’t thought of, querying the stack frequently, aggregating and massaging the results, and spitting out its models. This thing takes WMI and does real math against the results, math and pivots even an Excel jockey could appreciate. One of the VMTurbo product managers I worked with told me that they’d like to use Powershell, but powershell queries were still to slow whereas WMI could be queried rapidly.
It produces great reports I could never quite build in SCOM: By the end of day two, I had PDFs on CPU, Storage & network bandwidth consumption, top consumers, projections, and a good sense of current state vs desired state. Of course you can automate report creation and deliver via email etc. In the old days it was hard to get simple reports on CSV space free/space used; VMTurbo needed no special configuration to see how much space was left in a CSV
Integrates with AD: Expected. No surprises.
It’s low impact: I gave the VM 3 CPU and 16GB of RAM. The .vhd was about 30 gigabytes. Unlike SCOM, no worries here about the Observer Effect (always loved it when SCOM & its disk-intensive SQL back-end would report high load on a LUN that, you guessed it, was attached to the SCOM VM).
A Eureka! style moment: A software developer I showed the product to immediately got the concept. Viewing infrastructure as a supply chain, the heat map showing current state and desired state, these were things immediately familiar to him, and as he builds software products for PE, I considered that good insight. VMTurbo may not be your traditional operations manager, but it can assist you in translating your infrastructure into terms & concepts the business understands intuitively.
I was comfortable with its recommendations: During #VFD3, there was some animated discussion around flipping the VMTurbo switch from a “Hey! Virtualization engineer, you should do this,” to a “VMTurbo Optimize Automagically!” mode. But after watching it for a few weeks, after putting the APM together, I watched its recommendations closely. Didn’t flip the switch but it’s there. And that’s cool.
You can set it against your employer’s month end schedule: Didn’t catch a lot of how to do this, but you can give VMTurbo context. If it’s the end of the month, maybe you’ll see increased utilization of your finance systems. You can model peaks and troughs in the business cycle and (I think) it will adjust recommendations accordingly ahead of time.
Cost: Getting sensitive here but I will say this: it wasn’t outrageous. It hit the budget we had. Cost is by socket. It was a doable figure. Purchase is up to my PE, but I think VMTurbo worked well for PE’s particular infrastructure and circumstances.
No sugar coating it here, this thing’s built for VMware: All vendors please take note. If VMware, nomenclature is “vCPU, vMem, vNIC, Datastore, vMotion” If Hyper-V, nomenclature is “VM CPU, VM Mem, VMNic, Cluster Shared Volume (or CSV), Live Migration.” Should be simple enough to change or give us 29%ers a toggle. Still works, but annoying to see Datastore everywhere.
Interface is all flash: It’s like Adobe barfed all over the user interface. Mostly hassle-free, but occasionally a change you expected to register on screen took a manual refresh to become visible. Minor complaint.
Doesn’t speak SMB 3.0 yet: A conversation with one product engineer more or less took the route it usually takes. “SMB 3? You mean CIFS?” Sigh. But not enough to scuttle the product for Hyper-V shops…yet. If they still don’t know what SMB 3 is in two years…well I do declare I’d be highly offended. For now, if they want to take Hyper-V seriously as their website says they do, VMTurbo should focus some dev efforts on SMB 3 as it’s a transformative file storage tech, a few steps beyond what NFS can do. EMC called it the future of storage!
Didn’t talk to my storage: There is visibility down to the platter from an APM perspective, but this wasn’t in scope for the trial we engaged in. Our filer had direct support, our Nimble, as a newer storage platform, did not. So IOPS weren’t part of the APM calculations, though free/used space was.
Trusted Install & taking ownership of reg keys is required: So remember how I said VMTurbo was agent-less, using WMI in an ingenious way to gather its data from VMs and hosts alike? Well, yeah, about that. For Hyper-V and Windows shops who are at all current (2012 or R2, as well as 2008 R2), this means provisioning a service account with sufficient permissions, taking ownership of two Reg keys away from Trusted Installer (a very important ‘user’) in HKLMCLSID and one further down in WOW64, and assigning full control permissions to the service account on the reg key. This was painful for me, no doubt, and I hesitated for a good week. In the end, Trusted Installer still keeps full-control, so it’s a benign change, and I think payoff is worth it. A Senior VMTurbo product engineer told me VMTurbo is working with Microsoft to query WMI without making the customer modify the registry, but as of now, this is required. And the Group Policy I built to do this for me didn’t work entirely. On 2008 R2 VMs, you only have to modify the one CLSID key
Soup to nuts, I left PE pretty impressed with VMTurbo. I’m not joking when I say it probably could optimize my former virtualized environment better than I could. And it can do it around the clock, unlike me, even when I’m jacked up on 5 Hour Energy or a triple-shot espresso with house music on in the background.
Stepping back and thinking of the concept here and divesting myself from the pain of install in a Hyper-V context: products like this are the future of IT. VMTurbo is awesome and unique in an on-prem context as it bridges the gap between cost & operations, but it’s also kind of a window into our future as IT pros
That’s because if your employer is cloud-focused at all, the infrastructure-as-market-economy model is going to be in your future, like it or not. Cloud compute/storage/network, to a large extent, is all about supply, demand, consumption, production and bursting of resources against your OpEx budget.
What’s neat about VMTurbo is not just that it’s going to help you get the most out of the CapEx you spent on your gear, but also that it helps you shift your thinking a bit, away from up/down, latency, and login times to a rationalized economic model you’ll need in the years ahead.
Greetings to you Labworks readers, consumers, and conversationalists. Welcome to the last verse of Labworks Chapter 1, which has been all about building a durable and performance-oriented ZFS storage array for Hyper-V and/or VMware.
Today we’re going to circle back to the very end of Labworks 1:1, where I assigned myself some homework: find out why my writes suck so bad. We’re going to talk about a man named ZIL and his sidekick the SLOG and then we’re going to check out some Excel charts and finish by considering ZFS’ sync models.
But first, some housekeeping: SAN2, the ZFS box, has undergone minor modification. You can find the current array setup below. Also, I have a new switch in the Daisetta Lab, and as switching is intimately tied to storage networking & performance, it’s important I detail a little bit about it.
Labworks 1:4 – Small Business SG300 vs Catalyst 2960S
Cisco’s SG-300 & SG-500 series switches are getting some pretty good reviews, especially in a home lab context. I’ve got an SG-300 and really like it as it offers a solid spectrum of switching options at Layer 2 as well as a nice Layer 3-lite mode all for a tick under $200. It even has a real web-interface if your CLI-shy, which
I’m not but some folks are.
Sadly for me & the Daisetta Lab, I need more ports than my little SG-300 has to offer. So I’ve removed it from my rack and swapped it for a 2960S-48TS-L from the office, but not just any 2960S.
No, I have spiritual & emotional ties to this 2960s, this exact one. It’s the same 2960s I used in my January storage bakeoff of a Nimble array, the same 2960s on which I broke my Hyper-V & VMware cherry in those painful early days of virtualization, yes, this five year old switch is now in my lab:
Sure it’s not a storage switch, in fact it’s meant for IDFs and end-users and if the guys on that great storage networking podcast from a few weeks back knew I was using this as a storage switch, I’d be finished in this industry for good.
But I love this switch and I’m glad its at the top of my rack. I saved 1U, the energy costs of this switch vs two smaller ones are probably a wash, and though I lost Layer 3 Lite, I gained so much more: 48 x 1GbE ports and full LAN-licensed Cisco IOS v 15.2, which, agnostic computing goals aside for a moment, just feels so right and so good.
And with the increased amount of full-featured switch ports available to me, I’ve now got LACP teams of three on agnostic_node_1 & 2, jumbo frames from end to end, and the same VLAN layout.
Here’s the updated Labworks schematic and the disk layout for SAN2:
Disk Type, Quantity, Size, Format, Speed, Function
WD Red 2.5″ with NASWARE, 6, 1TB, 4KB AF, SATA 3 5400RPM, Zpool Members
Samsung 840 EVO SSD, 1, 128GB, 512byte, SATA 3, L2ARC Read Cache
Labworks 1:5 – A Man named ZIL and his sidekick, the SLOG
Labworks 1:1 was all about building durable & performance-oriented storage for Hyper-V & VMware. And one of the unresolved questions I aimed to solve out of that post was my poor write performance.
Review the hardware table and you’ll feel like I felt. I got me some SSD and some RAM, I provisioned a ZIL so write-cache that inbound IO already ZFS, amiright? Show me the IOPSMoney Jerry!
Well, about that. I mischaracterized the ZIL and I apologize to readers for the error. Let’s just get this out of the way: The ZFS Intent Log (ZIL) is not a write-cache device as I implied in Labworks 1:1.
The ZIL, whether spread out among your rotational disks by ZFS design, or applied to a Separate Log Device (a SLOG), is simply a synchronous writes mechanism, a log designed to ensure data integrity and report (IO ACK) back to the application layer that writes are safe somewhere on your rotational media. The ZIL & SLOG are also a disaster recovery mechanisms/devices ; in the event of power-loss, the ZIL, or the ZIL functioning on a SLOG device, will ensure that the writes it logged prior to the event are written to your spinners when your disks are back online.
Now there seem to be some differences in how the various implementations of ZFS look at the ZIL/SLOG mechanism.
Nexenta Community Edition, based off Illumos which is the open source descendant of Sun’s Solaris, says your SLOG should just be a write-optimized SSD, but even that’s more best practice than hard & fast requirement. Nexenta touts the ZIL/SLOG as a performance multiplier, and their excellent documentation has helpful charts and graphics reinforcing that.
In contrast, the most popular FreeBSD ZFS implementations documentation paints the ZIL as likely more trouble than its worth. FreeNAS actively discourages you from provisioning a SLOG unless it’s enterprise-grade, accurately pointing out that the ZIL & a SLOG device aren’t write-cache and probably won’t make your writes faster anyway, unless you’re NFS-focused (which I’m proudly, defiantly even, not) or operating a large database at scale.
What’s to account for the difference in documentation & best practice guides? I’m not sure; some of it’s probably related to *BSD vs Illumos implementations of ZFS, some of it’s probably related to different audiences & users of the free tier of these storage systems.
The question for us here is this: Will you benefit from provisioning a SLOG device if you build a ZFS box for Hyper-V and VMWare storage for iSCSI?
I hate sounding like a waffling storage VAR here, but I will: it depends. I’ve run both Nexenta and NAS4Free; when I ran Nexenta, I saw my SLOG being used during random & synchronous write operations. In NAS4Free, the SSD I had dedicated as a SLOG never showed any activity in zfs-stats, gstat or any other IO disk tool I could find.
One could spend weeks of valuable lab time verifying under which conditions a dedicated SLOG device adds performance to your storage array, but I decided to cut bait. Check out some of the links at the bottom for more color on this, but in the meantime, let me leave you with this advice: if you have $80 to spend on your FreeBSD-based ZFS storage, buy an extra 8GB of RAM rather than a tiny, used SLC or MLC device to function as your SLOG. You will almost certainly get more performance out of a larger ARC than by dedicating a disk as your SLOG.
Labworks 1:6 – Great…so, again, why do my writes suck?
Recall this SQLIO test from Labworks 1:1:
As you can see, read or write, I was hitting a wall at around 235-240 megabytes per second during much of “Short Test”, which is pretty close to the theoretical limit of an LACP team with two GigE NICs.
But as I said above, we don’t have that limit anymore. Whereas there were once 2x1GbE Teams, there are now 3x1GbE. Let’s see what the same test on the same 4KB block/4KB NTFS volume yields now.
SQLIO short test, take two, sort by Random vs Sequential writes & reads:
By jove, what’s going on here? This graph was built off the same SQLIO recipe, but looks completely different than Labworks 1. For one, the writes look much better, and reads look much worse. Yet step back and the patterns are largely the same.
It’s data like this that makes benchmarking, validating & ultimately purchasing storage so tricky. Some would argue with my reliance on SQLIO and those arguments have merit, but I feel SQLIO, which is easy to script/run and automate, can give you some valuable hints into the characteristics of an array you’re considering.
Let’s look at the writes question specifically.
Am I really writing 350MB/s to SAN2?
On the one hand, everything I’m looking at says YES: I am a Storage God and I have achieved #StorageGlory inside the humble Daisetta Lab HQ on consumer-level hardware:
SAN2 is showing about 115MB/s to each Broadcom interface during the 32KB & 64KB samples
Agnostic_Node_1 perfmon shows about the same amount of traffic eggressing the three vEthernet adapters
The 2960S is reflecting all that traffic; I’m definitely pushing about 350 megabytes per second to SAN2; interface port channel 3 shows TX load at 219 out of 255 and maxing out my LACP team
On the other hand, I am just an IT Mortal and something bothers:
CPU is very high on SAN2 during the 32KB & 64KB runs…so busy it seems like the little AMD CPU is responsible for some of the good performance marks
While I’m a fan of the itsy-bitsy 2.5″ Western Digitial RED 1TB drives in SAN2, under no theoretical IOPS model is it likely that six of them, in RAIDZ-2 (RAID 6 equivalent) can achieve 5,000-10,000 IOPS under traditional storage principles. Each drive by itself is capable of only 75-90 IOPS
If something is too good to be true, it probably is
Sr. Storage Engineer Neo feels really frustrated at this point; he can’t figure out why his writes suck, or even if they suck, and so he wanders up to the Oracle to get her take on the situation and comes across this strange Buddha Storage kid.
Labworks 1:7 – The Essence of ZFS & New Storage model
In effect, what we see here is is just a sample of the technology & techniques that have been disrupting the storage market for several years now: compression & caching multiply performance of storage systems beyond what they should be capable of, in certain scenarios.
As the chart above shows, the test2 volume is compressed by SAN2 using lzjb. On top of that, we’ve got the ZFS ARC, L2ARC, and the ZIL in the mix. And then, to make things even more complicated, we have some sync policies ZFS allows us to toggle. They look like this:
The sync toggle documentation is out there and you should understand it it is crucial to understanding ZFS, but I want to demonstrate the choices as well.
I’ve got three choices + the compression options. Which one of these combinations is going to give me the best performance & durability for my Hyper-V VMs?
SQLIO Short Test Runs 3-6, all PivotTabled up for your enjoyment and ease of digestion:
As is usually the case in storage, IT, and hell, life in general, there are no free lunches here people. This graph tells you what you already know in your heart: the safest storage policy in ZFS-land (Always Sync, that is to say, commit writes to the rotationals post haste as if it was the last day on earth) is also the slowest. Nearly 20 seconds of latency as I force ZFS to commit everything I send it immediately (vs flush it later), which it struggles to do at a measly average speed of 4.4 megabytes/second.
Compression-wise, I thought I’d see a big difference between the various compression schemes, but I don’t. Lzgb, lz4, and the ultra-space-saving/high-cpu-cost gzip-9 all turn in about equal results from an IOPS & performance perspective. It’s almost a wash, really, and that’s likely because of the predictable nature of the IO SQLIO is generating.
Last point: ZFS, as Chris Wahl pointed out, is a sort of virtualization layer atop your storage. Now if you’re a virtualization guy like me or Wahl, that’s easy to grasp; Windows 2012 R2’s Storage Spaces concept is similar in function.
But sometimes in virtualization, it’s good to peel away the abstraction onion and watch what that looks like in practice. ZFS has a number of tools and monitors that look at your Zpool IO, but to really see how ZFS works, I advise you to run gstat. GStat shows what your disks are doing and if you’re carefully setting up your environment, you ought to be able to see the effects of your settings on each individual spindle.
Now look at this gstat sample. Under SQLIO-load, the zvol is showing 10,000 IOPS, 300+MB/s. But ada0-5, the physical drives, aren’t doing squat for several seconds at a time as SAN2 absorbs & processes all the IO coming at it.
That, friends, is the essence of ZFS.
Links/Knowledge/Required Reading Used in this Post: