Re: Hyper Converged newness vs Big Iron oldness
OoooOOOooooo, Good topic....
7037 posts • joined 31 May 2010
OoooOOOooooo, Good topic....
If my phone 'works as it always did" then clearly the phone network's gotten worse. Damned thing drops calls 10 times a day. Never used to have that problem.
What is important, I think, is not "how many files can you manage", but "how many files can be managed with acceptable performance. With traditional file systems (extX, NTFS, etc) this really does drop off dramatically after about 10M files. After that, adding spindles and controller cards doesn't matter. The issue is the file system, it's complexity, it's size, how much of it fits into RAM and other such considerations.
It's the tipping point where "whole system" concerns become far greater than spindle throughput.
So it's worth a right proper "try before you buy" with file systems. Get the wrong one and you could wind up with a filer that's dog slow and no amount of added disks will make it faster.
I am not talking about the ability to store the files, but rather the approximate point at which file access performance starts to degrade significantly. With NTFS this absolutely is 10M files. With ReFS, you can push that a little higher. ZFS, EXT3 and EXT4 all start to see some pretty big drop offs between 10M and 15M files. NetApp's more recent files (past three years or so) are some of the better setups; I've seen 30M files before the access starts to degrade.
Note: you can put more files on these file systems. That doesn't mean that doing so will allow accessing files on a 100M file-populated system is anywhere near as fast as accessing a file on a 10M file-populated system. Especially if the system stores these files for more than just cold archival.
If these files are accessed even semi-regularly. (Let's say 2M files a day are read or written of 10M) then the system is spending all it's time faffing about with the index. The more files you touch, the more time the system cranks away on the index rather than the data blocks.
Different file systems handle it differently, but the rule of thumb is that you have to start paying very close attention to your file server designs once you surpass 10M files. (Dramatically upping RAM, for example, or considering using a file system that can offload the index to SSD, etc...)
That is a good read. Thanks for the link!
Is it "in a mobile device"?
You live in a fairy land.
"It might make a useful article, how to chose & set up a router and NAS + few machines so you can VPN in and access your data or desktop with tolerable risk?"
Problem = solved.
"Make NAS visible to outside world".
Um hmm. And do you want to to wager how long it would take me to crack your off-the-shelf NAS with 13 pending updates and 50 known vulns...only 35 of which the vendor has patched?
Hell even I have black vulns for most of the major NASes. DO NOT EXPOSE NASES TO THE INTERNET. EVER. DO NOT DO.
If CSIS or CESC dig into my data, they must answer to a judge. Or, if not, then at the very least I cannot be sued in a Canadian court for their misdeeds.
If I am complying with Canadian law and our security services decide to break that law, my clients cannot win in court against me: they have to fight the spooks.
Ah, but if I host my data elsewhere...can I really say that I am obeying Canadian law? The data is not subject only to Canadian law if it's outside Canada. What's more, it may be impossible to know whose laws it is subject to (large cloud providers move data around the globe as part of routine maintenance).
So, unless I have a crackerjack legal team, why take the risk? Keep my data in Canada, let my customers duke privacy concerns out with our government.
American legal attack surface > 0 = bad.
But this is the human genome that's not on a mobile device!
So..."any tomfoolery that makes Microsoft look dominant is good, any tomfoolery that makes someone else look dominant is bad."
Glad to see where your biases lie.
Does that count things like Office 365 as "cloud"? Seems to me Amazon doesn't have a whole lot of Amazon SaaS services. They do IaaS and PaaS. The SaaS side of things is provided by Amazon's customers.
If Microsoft is incorporating SaaS applications like Office 365 into their cloud revenues in order to get this "Amazon-beating" figure, is this factoring in the amount of revenue lost by the traditional Office, Exchange, Dynamics, etc groups?
And shouldn't you then count Amazon's tat bazzar as "cloud revenue"? If Microsoft is including revenue other than IaaS and PaaS, Amazon should include "Physical Items as a Service" in their cloud reporting. Does Amazon report it's Content as a Service as part of cloud?
It's really hard to know if things are being compared like for like here.
Because PowerShell isn't intuitive, and there exist alternatives. Period.
If Microsoft were to do that they would lose their customer base instantly.
"More than 30% of all ODM output is purchased by the large cloud vendors. Anybody still doing premise installations is a small customer in comparison."
Sure, but 70% of all ODM output is on premises. It's not like cloud displaced on premesis entirely. It became a supplement; CAGR for the past few years was in the cloud, but on premises purchases have stayed pretty flat.
"You have this backwards. Customers stampeded toward the cloud and left Microsoft standing in the field."
No. Wrong. Absolutely wrong. Cloud adoption is a supplement to on premises purchasing. It is not a displacement. Clouds are being used tactically, for specific workloads that work well in teh cloud. They are not a bulk replacement for all workloads, and only a handful of businesses are treating them as such. Almost every single startup that starts in the cloud goes hybrid/on-premises within 18 months simply due to cost.
"System Center and VMM users are, it seems, considered more deserving of attention by Microsoft."
Funny how actually having competition that's notably superior can motivate even Microsoft to deign to add minor additions. A few more years of this and they might even stop behaving as though they are an inevitable monopoly!
The big question yet to be resolved is how Microsoft is going to react to the changes that have been happening of late. "Cloud first, mobile first, customers/partners/staff last" has succeeded in it's original goal: drive cloud adoption. The cost of this, however, has been a gut punch to on premises revenues in all segments.
Where the lines cross a little is that Microsoft is losing customers. There are a significant number of companies that simply don't want to go cloud, and for all the critical markets (except traditional desktop OSes) there are viable alternatives. In addition, some of those convinced to go cloud are choosing vendors other than Microsoft for their cloud services.
In and of itself neither of these two things is negative. Not all customers are equal, and it remains to be seen if Microsoft's approach is driving away the valuable customers or the burdensome ones.
At some point, however, there will be a massive shift. Unless Microsoft changes it's approach to on premises customers the decline will accelerate. As it does, Microsoft will reach a point where merely treating on premises users as second class citizens isn't enough, and they simply retire the whole concept.
When will that occur? What will the revenue fallout of that look like? And in the intervening months/years will Microsoft successfully convert the holdouts to their American controlled public infrastructure?
Microsoft used to be purchased by companies for two reasons: 1) Ease of use and 2) predictability. They were a stable, predictable, comfortable option. Sure, they were more expensive, but you know what they were up to and surprises were few.
That isn't the Microsoft of today. There is a constant feeling that on premises customers are one bad quarter away from getting PlaysForSured. That's before we get into discussions about development stacks and the ever shifting sands underneath them.
Microsoft are smart, capable and make products that are - for the most part - competitive with what's on offer from others. But the existing strategy of herding customers towards the cloud had led to some pretty bitter alienation and it is starting to show in revenue.
Can Microsoft continue to be reactionary; doling out only the bare minimum of updates (such as what has been presented for SCVMM on premises customers in this article)? Will playing silly buggers with products where they think they have a monopoly still be a viable approach as Silicon Valley's startup engine roars ever louder, and Microsoft's customers eye its empire ever more hungrily?
Or will Microsoft's arrogance and belief that it can control its customer base lead to a perpetual negative CAGR for the company as a whole and a slow fade to being nothing more than an American public cloud provider, with a politics-limited TAM and an increasing inability to differentiate itself from other American public cloud providers?
Can Microsoft win on merit with it's current approach? If not, what changes are required for it to do so? If there are changes required, can the body corporate make them, or will Microsoft fall to its own corporate hubris?
I am no longer certain Microsoft has mastery over the art of pricing the cost of staying loyal below the cost of moving away, and I am absolutely positive that the desire to move away is mounting. Interesting times. 15 years ago, who would have thought this day would ever have arrived?
"Sounds like you have not learned the difference yet. Hyper-V server is not ain any way crippled. It's fully functional with all the Hyper-V features but with lower resource requirements and reduced attack surface."
All the Hyper-V features, sure. But not the other features that make Windows Server such a great operating system, nor with a usable GUI. But hey, if you focus your thoughts narrowly enough you can prove any point, eh?
"Only if you are incompetent. We run nearly all our Windows Servers and Hyper-V servers without a GUI, unless required for a specific third party application. As everyone should do. Hence why no GUI is the default when you install Windows Server."
So the overwhelming majority of humans who are visually oriented are "incompetent"? Even those who can manage things via command line but prefer easier, simpler, faster and more intuitive interfaces? They're "incompetent" because they don't share your mutation that allows you to bypass billions of years of visual reference response wiring?
Jibbers fucking Crabst. you're arrogant.
"All the management tools can be run remotely from your desktop, if you can't cope with Powershell. The tools are identical to if you had a GUI on the server."
Again, you completely ignore what I actually wrote. These tools require you to use an operating system that has been overwhelmingly rejected by the world, and they are significantly inferior to the tools offered by other hypervisors.
All you are doing here is spouting Microsoft marketing propaganda. Again. Why do you feel the need to do this? Don't you have friends? A family? Or even an overused prono mag that occupies your attention such that you aren't constantly online haranguing people with Microsoft's broken vision for solving problems of 5 years ago?
Microsoft isn't the solution to all problems. Their approach to management interfaces, UIs, or even how they treat their customers, partners and staff sure as hell aren't great. They are one possible solution to today's problems, with a passable hypervisor and shitty management tools. They also have a fetish for subscription licensing and getting my data into an insecure country run by madmen.
And even if they somehow, magically, were to become the ultimate solution to all ills, your hostile and arrogant approach to "selling" people on the benefits of your tribal brand is completely fucking cracked. At this point, I wouldn't care if Microsoft handed out unicorns and bags of cash with every copy of Hyper-V, I wouldn't use the goddamned thing because it would mean that my user community was people like you!
We get it. You really love Microsoft. Congratulations. Do you want a statue or something? The rest of us aren't fucking robots,ingesting integrals and pissing derivatives. We don't want to sit in a dark room and string together scripts to run our infrastructure. We don't even consider "running infrastructure" to be a full time job.
The rest of the world has too many important things to do to waste our time becoming "specialists" in some bit of infrastructure. We have to go forth and make the company money by doing things like working out how to integrate and automate that infrastructure so that we can make more money with less manpower, or make more widgets with fewer mistakes.
To do that, we rely on intuitive user interfaces. Things that allow us to not touch some element of the infrastructure for months - even years - but figure it out when needed in a matter of seconds. No memorizing commands. No scripts in a folder buried in some hierarchy. No debugging piles of linked PowerShell.
We have money to make, things to do, lives to live. And we don't want to spend it dicking around setting up 8 fucking servers just to get a basic environment working, or "retraining" every 18 months to learn another college semester's worth of commands just to manage something that should have been easily handled by the goddamned developers in their UI to start with.
All the other major companies playing this game seem able to crank out usable interfaces. Even the bloody open source projects. But Microsoft can't. And they don't care enough to even try. I mean hell, Microsoft have even killed Technet because they give zero fucks about hobbiests, junior admins or those who don't have enterprise resources backing their expensive and perpetual ongoing training.
You advocate a world where systems administration is a massively specialized and specialist-focused endeavor and you champion the vendor that still perpetuates this myth. Everyone else wants you to be able to learn the basics and then adapt smoothly as new technologies come out.
If for no other reason than sheer self preservation it makes sense not to engage with Microsoft. Because Microsoft wants to make such specialists that we all know more and more about less and less until we all know absolutely everything there is about absolutely nothing.
And when our area of expertise is rolled into their cloud service and no longer available locally? Too bad, so sad. Reskill. At enormous cost in time and money.
Why, why, why, WHY would anyone choose that? Why would they choose this archaic approach to things when superior alternatives exist? Microsoft isn't faster. It's not more efficient. It's not "better" in any objective sense.
Microsoft is merely one competitor amongst many, and if they want to win the hearts, minds and wallets of the many then they have to offer a better experience - in the short term and the long term - than their competitors.
They have flat out fucking failed at this for some time, an nowhere is that more clear than with hypervisors and their management tools.
Maybe if you are ever able to comprehend the above you'll understand why so very many have left Microsoft's orbit, and why oh so very many are never going back.
"UX != GUI"
Certainly not for those who keep their pee in jars, nope. But a substantial percentage of the human race is visually oriented. So let's agree to disagree, hmm? There's the niche of people who do quadratic equations in their head for fun, then there's the entire rest of our species. For the overwhelming majority UX = GUI.
I haven't seen much OpenVZ in production, nor heard of it used in more than some niche cases. It's out there, to be sure, but like all container tech it's not yet ascendant. Try this series of 4 articles for my look at containers, including OpenVZ:
@Getmo Ah, but don't forget the first rule of brand tribalism: compare the latest, greatest from your favourite brand to something several years out of date by the competition. Don't do things like "compare the most deployed version of your favourite brand versus the most deployed version of the competition", or even newest to newest. Cherry pick!
The think about Mist.io is that the whole "it's in the cloud" is a huge seller. It would be really cool if you could set up partnerships with companies in different geos and license the full version out to them. They could then run a Mist.io cloud service for geolocal customers and still be within their legal framework. Those customers get the full advantage of the Mist.io cloud offering without the lengthy setup.
Now THAT arrangement...that would be worth some damned good money, because Mist.io is a great product. :)
I don't recall saying Hyper-V server is "part of Windows Core". I recall saying the difference between Windows Core + Hyper-V and Hyper-V server is functionally non existent, so long as you leave Windows Core at a minimal install.
Except, you know, the part where Windows Core costs money, isn't completely fucking WONTFIX broken and can be turned into a real server with a few commands. But if you're doing everything remotely, then the difference between the two is "licensing" and "Hyper-V server is slightly more miserable to get connected and set up than Windows Core".
As for "learn some PowerShell", I know lots of PowerShell. Knowing PowerShell and liking it are completely different things. I also am aware of how people actually use their tools in the real world. Survey after survey, study after study.
You'll find "people who use PowerShell for day to day management" are a pretty small subset. most administrators are just fine using PowerShell for automation. Only the "pee in jars" types use it for management.
And to be perfectly blunt, if I was going to go full "pee in jars" on something, why the metric fnord wouldn't I just go full Openstack? It doesn't cost me my left testicle, and is so modular and open I don't get locked into anything. Openstack is superior in virtually every way to Microsoft and their wretched ecosystem.
And holy pants batman, they even recognize that real human being need proper GUIs to work with tools on a day to day basis without going completely batshit bananas bonkers. That's right, ladies and germs: it's the open source community are the ones who have embraced ease of use and Microsoft and their fanboys are the ones telling us to manage every little thing from a command line.
So yeah, take your brand tribalism and go right back under your bridge. You can come back out when Microsoft gives enough fucks about their community to bring back Technet subscriptions and not a moment before.
Just buy a subscription from Azure! What do you mean you don't want to pay every month forever? And what do you mean "data sovereignty" and "foreigner rights"? Now you're just being difficult.
"Considering how bloody awful VMware's GUI tools actually are, for that statement to be true then Microsoft's must truly be beyond abysmal."
They are. Read the whole bit about the fustercluck you have to go through just to mount an ISO to install a VM. Microsoft gives negative fucks about their user base. If Windows 8 didn't teach you that, well...nothing will.
"AWS is a direct competitor"
Fucking Americans. there ate 7 billion of us that don't live inside your Orwellian nightmare and who don't want to. Your yankee-run public clouds aren't a competitor at all. Not for us. If we're going to use cloud computing it needs to be from companies run in countries that acknowledge that we have rights. And no, a foreign datacenter isn't enough. Ask the Irish.
"Limited PoS that you have to CODE SCRIPTS!!! in RUBY!!! to get it to do what you want. Utterly inferior in pretty much every way to SCCM and Opalis. We binned it long ago and never looked back."
So you don't actually know at all what you're talking about. Congratulations. You have proven your ignorance. Now why don't you attach your name to your words and own them, or are you the coward your tickbox says you are?
"Versus Hyper-V Server:
Install it - and it's secure and auto updated by default."
From this you are implying that Linux is insecure and does not auto-update by default. You're wrong.
CentOS is secure by default. You have to add things to it if you want to more easily manage it, or you want to make it more usable whilst staying secure. This is no different than Hyper-V.
Hyper-V needs to be joined to a domain controller to be actually usable. Then it gets it's config from the domain controller. CentOS needs the puppet agent installed, then it pulls down it's config from puppet. No real difference.
What is different is that CentOS isn't completely fucking broken if I need to work on the node individually without the primary management tools. Hyper-V is.
As virtually nobody uses Hyper-V server - stats are very clear on this, they use Windows Server - both OSes can grow from a "core" install containing not bloody much of anything to complete OSes capable of doing anything. By default, they install very small, very secure and with not bloody much to them.
Oh, and I'll take my unpatched CentOS with SELINUX on by default over the unpatched Windows any day. There's a gap between "installed" and "fully up to date" while the systems are doing their update thing, and Linux is a hell of a lot less likely to be pwned during that time. (Always assume your network is compromised!)
Meanwhile, I only have to reboot once after install with CentOS. Puppet goes on, configs it all, updates download, it reboots once and it's ready to rock. Secure the whole time, and actually usable, from console, remotely, what-have-you.
And I don't need a domain controller! I'm not locked in to an ecosystem by one vendor that will just turn the knobs on pricing whenever they feel like it! In fact, I don't even have to pay for licences for my production systems at all. I only need support for my test & dev systems, where we'll actually be prototyping workloads. I can run entire datacenters off CentOS (or derivatives) as the production distro for free...just like tens of thousands of other companies.
And no, there's no real difference between the "free" version and the "support included" version. The delta between Microsoft's "free" Hyper-V server and the full Windows Server is so big that Hyper-V server's adoption is limited to a handful of hyper-scale deployments and some real die hard Microsoft fans.
But if you know Redhat then you know CentOS. They're the same goddamned thing. For that matter, there difference between SuSE and RedHat isn't that big anymore, and the skills cross over (and the default configs are so close) that it is less of a jump than going from Server 2003 to Server 2008 R2.
Microsoft has a great hypervisor. But they've fucked up every aspect of the ecosystem, from pricing and limitations to management tools to chasing away all their partners in the name of getting their hostages moved to Azure.
Put your fanboy goggles down and actually look at the bloody ecosystems as a whole, will you? It isn't about "company A has this feature" or "company B can do this, if you beat them properly". It's about the whole of the experience. And it's about the experience over years.
How is actually setting up an environment from scratch? How is maintianing that environment over time? What are the long term costs? The time commitments? The training requirements? What is lock-in like? How does the vendor - and/or the community - treat customers/users/administrators?
Are they trying to push you away from running your own infrastructure and towards paying even more to run your workloads on their American public cloud where they're beholden to a sociopathic government that doesn't' recognize the basic rights of it's own citizens, let alone the citizens of other countries?
If I pick a hypervisor ecosystem what can I expect for myself (as the administrator) and my company over the next decade? The next two?
Right from the installation of the "free" server through to "how you are treated over the course of decades" Microsoft is consistently "only just acceptable enough to not get binned, if you are already locked in to them." "They do the bare minimum and not one iota more" is not a great recommendation. It works if you're a monopoly, but it's worth nothing where there's actual competition.
Microsoft will never learn that lesson. I can only pray VMware does. In the meantime, there are alternatives, and they're damned fine alternatives. Finally. they took long enough to get that way.
Microsoft has a WONTFIX-class broken freebie version of Hyper-V server that is nothing more than a crippled version of Server Core + Hyper-V role. Congratulations. Free ESXi is *still* better, and KVM is a better choice than both. OOooooOOooOOooooo.
Of course, nobody actually uses it. If you look at the adoption stats, much to my great dismay, virtually everyone using Hyper-V as their hypervisor is doing so not only with full versions of Windows Server...they do it with the GUI installed.
Now, why might that be? It's because of the management tools, and because they want ease of setup. Management tools are important. I know that's something you are terrified to discuss - because it's where Microsoft looks like der lumpenfools - but it's true. Real people and real sysadmins in the real world actually give fucks about ease of use.
Hyper-V server is broken. Server Core + Hyper-V is a miserable bitch to work with if something's gone wrong. So virtually everyone deploys full fat Windows as a hypervisor. Study after study after study finds this.
What's interesting is that in terms of total deployed nodes, hyper-v server manages to almost equal Windows Server /w GUI deployments because a handful of really large deployments use it.
Now, what's more interesting is that despite installing the full-GUI version of Windows Server, virtually nobody deploying more tan two nodes uses the Windows Server native Hyper-V tools. Hyper-V in practice for virtually every deployment means the full fat GUI to run the nodes and SCVMM to run the clusters.
While 5nine has cult-like status amongst the SMB market they have only recently made any sort of break through into the midmarket and they are functionally absent from the enterprise. This means that small clusters can be - and increasingly are - run by 5nine, but large (or multiple) clusters are run by SCVMM.
So when discussing Hyper-V in any practical sense you have the following:
1) Full fat Windows installed in order to run VMs and have an actual GUI to configure the underlying hypervisor.
2) 5Nine for small deployments.
3) A drop-and-cut upgrade to SCVMM after you reach the magic pumpkin point with 5Nine.
More critically, 5Nine's uptake has been dramatic in the past, oh 18 months or so. The last 8ish months have in really been the part where the name has spread and it is discussed openly. (Which is critical, because before about 8 months ago people treated using 5Nine as a point of shame.)
What's not happening is people installing the broken Hyper-V server, fighting with each node and then using Windows 8 to remotely manage it with Microsoft's frustrating remote management tools. (Which is the other part of this that you always ignore. Only truly ill people want to use Windows 8 for anything, which has really suppressed the uptake of Hyper-V server.)
So a discussion about Hyper-V is a discussion about Windows Server. Like it or not. Cheers.
RIght. Barkeep, drinks all around. And make sure they're flammable, damn it!
You are correct, I am extremely biased towards ease of use and lack of frustration when managing a product. And the delta between some of the experiences here is far more than "a little bit better". It is most definitely into "completely fucking wrecks it" territory.
Also: if you speak "the Queen's English" you should understand the use of "wrecks" in this context. Unless you're from that bizarre island of poncy gits with sticks up their pigus. Then you don't so much "speak" the language as snark it. That's okay, from down here all we notice from beyond your long noses is the disturbing amount of hair in them.
I know Puppet not only comes close to Windows management functionality, it surpasses it in many ways. The fact that you don't demonstrates nothing more than that you haven't actually learned anything about that which you are deriding.
Her Majesty's Canadian English, yes sir. Why, what bastard variant do you speak?
yum update -y added to cron
Webmin for firewall config and changing default ports
For the love of Jibbers *anything* other than Microsoft's CA
Oh look, more secure than 90% of what's out there in 10 minutes. *shrug*
Always known the difference. The thing you never understood is that the difference doesn't matter. It's the management tools that matter.
But since you sup on marketing, not infrastructure management, I expect you don't understand that yet.
Except these hold only replicas while primary workloads are served off the systems running those workloads. Big difference.
I don't even understand how you can run a modern Windows on a 15GB C:\ drive. 40GB VDI instances floweth over in 8 months. 75GB instances only go about 3 years before needing WinSXS purged. Server instances seem to be in the "80-100GB OS disk" range for usable installs of 3+years.
I sort of get the guy who can use 128GB disks on his endpoints and force users to keep data on the server, because he's augmenting the core storage with centralized storage. Even then, I'd look a little crosseyed at 128GB SSDs today, given that 240GB drives are the sweet spot pricing-wise, and Windows 10 looks to have the WinSXS problem in spades.
Why not buy a 240GB and "short stroke" it? Only using half the space means that the wear leveling algorithm has more cells to play with, making doubly sure you don't hit write life limits. And if you do need more space in the future it's not a rip and replace affair to get it.
As for the "30GB SSDs into PoS terminals", to me that falls under "embedded devices" rather than desktops. Single purpose devices that are probably never going to be updated and hang around until the end of time contributing to our IoT 33Billion device attack surface. Huzzah!
Seriously though, I've deployed a fair amount of PoS terminals, thin clients, etc...and even on those - even for Linux - I find 30GB is constraining. And if it's work to squeeze the OS and all relevant defenses in there today, 6 or 10 years from now space could be a real problem. Unless, of course, updates and maintenance aren't part of the plan.
Who or what is "the Fast Show". I tried checking it out on YouTube, but if this was supposed to be funny it completely bypassed me.
Number of PCs shipped != number of PC gamers.
I haven't worked for anyone with that little financial sense in quite some time. Trying to run Windows Server on a 30GB SSD with a projected install time greater than 6 years is called "being penny wise and pound foolish".
Examine the total cost of ownership over the life of the unit. This includes the cost of maintenance, downtime for upgrades, spares, electricity, bandwidth and so forth. I have spend my entire career as an SMB admin for the cheapest people alive, and I promise you that the sort of nonsense you advocate in that regard is far more costly over the life of the unit than simply buying a "sweet spot" drive and letting your OS grow.
I know it's very hard for some people to factor in the cost of manpower. THey think that being on salary makes their time cost nothing. For sysadmins this was true 15 years ago when there wasn't such a diversity of products to support. Now, the sheer volume of different kinds of hardware, software, networking etc that even the smallest of SMBs must support strains the ongoing maintenance capabilities of even the most strongly "pee in jars" sysadmin.
Try to make what you deploy as "fire and forget" as possible. That will require frontloading a few extra % in terms of hardware cost in order to recoup hundreds of % in operating costs.
Your entire insane rant falls apart right at the very beginning:
"Nutanix is lock-in simply because it's a proprietary stove-piped solution. Your data sits in a silo in a data format inaccessible by tools and protocols other than what the vendor allows you"
Nutanix uses the same formats the hypervisors use. There is nothing proprietary about the "data formats" at all. There is nothing preventing you from very easily moving from Nutanix off to someone else.
So basically, you're just making shit up. The question is...why?
The tears of enraged commenters power my happiness.
I rather like gardening, and I don't see why a lack of internet access would prevent me from obtaining the automation tools required to make farming much more than "gardening at scale". Or why a lack of internet access would make it harder or all that much more expensive to produce or ship those tools.
Actually 10Mb. Mega is capitalized. Bits are lower case, Bytes are upper case. Don't forget to distinguish between MB and MiB, as well.
Biting the hand that feeds IT © 1998–2017