Enterprise software is such a mystery to me. I can't imagine any company ran by <40 yr olds ever paying $100k+ a year to VMWare, Symantec, Oracle, ...
To me it just looks like older companies paying to keep from having to deal with Linux directly
To me it just looks like older companies paying to keep from having to deal with Linux directly
Those companies run all their boring, essential software (accounting, ERPs, etc.) on VMWare, Oracle, Azure, etc. Since software isn't their core business, they pay an external company to keep their software running. Like it or not, cloud providers are for software companies, and most companies have no interest in becoming a software company. They might have some small groups that do analytics or software in-house, and those groups might use cloud providers, but all the essential software will be run by a company you can call when something breaks, or if you're big enough embeds some staff in your office.
It's just good sense to outsource non-core functions. Software companies outsource hosting and datacenter stuff to AWS, etc. Why should enterprises deal with Linux directly?
I pay a barber to cut my hair. Sure, I could learn how to cut my own hair, but there's a good chance I'll spend a lot of time and still mess it up.
I pay a mechanic to fix my car for serious repair. Sure, I could learn how to replace a transmission, but there's a good chance I'll spend a lot of time and still mess it up.
If these (and the endless examples of paying professionals) are a tax for stupidity and laziness, then everyone is lazy and stupid. If you're not a software company, there's a good chance it's worth it to save the time to pay someone to set it up for you. Also, if Joe's Fish Market wants a POS system and needs a server for it and he decides to set it up himself, if he gets it up and running, what are the chances it'll stay up to day and not have a firewall issue? Probably pretty low compared to if a professional did it.
Joe wants to focus on selling fish. That's where he's a pro. I don't know anything about fish, but I could set up a server for him. We all have our own skills, and it's often worth while to leverage other people's skills, because they'll do it faster, better, and it will likely end up cheaper after time and security are taken into consideration.
Meaningless analogy. Here some other meaningless analogies:
Why would anyone ever own a frying pan? Not only do you need to rent a stove, you also need to hire a chef to cook with frying pan, you need a station with chefs that prep ingredients and so on. Just go to a restaurant.
You pay a mechanic to fix your car? Why would anyone ever own a car when you can rent a car whenever you need one.
You pay a barber to cut your hair? Why not just wear a wig!
And so on ...
For some — but not all — people, that’s a preferable option.
For some — but not all — companies, enterprise tech is a preferable option.
"Inevitable repairs after 4 years" is just a delusion, and silly.
Might I suggest you be honest with yourself and say you like the new car feel? Maybe even like it as a status symbol? Don't want to think about what tires to buy to replace worn-out ones?
I'm the opposite. I like having a scuffed-up car (or in this case, jeep), with dogs climbing all over the seats and sand everywhere from going off road. I'd rather use the thing and have fun on the trip.
Even if I drove my jeep off a cliff, called it a total loss, and got nothing back from insurance, I'd lose less money than you lost in 3 years of leasing -- and I've already had the thing longer than that.
Dead wrong.
I lease the cheapest car that is practical for my situation. I consider cars completely utilitarian tools. I get no joy from them, I hate driving and take public transit whenever I can.
I had very little money until I was in my late 20s when my career prospects finally took off. Cheap cars that failed frequently were a constant source of anxiety because I couldn’t get far enough ahead to maintain savings. When my career improved and I could afford small luxuries, I spent the money on leasing because I wanted to avoid the anxiety of cars breaking down. It was a gift to myself to lower my stress level.
EDIT: also, in northern Minnesota where the winters are hard and everything is covered in salt, car trouble DOES start after 4 years. Sometimes before.
Yes, they don't want to think about what tires to buy. Or whether it needs to be fixed or not. The fact that repairs are not in inevitability but a possibility is part of the point.
The parent is making the point that they're rather throw money at the problem and not have to think about it, not that they've precisely run the numbers on all of the possibilities and decided their outcome is the most cost-effective.
The OP prob doesn’t want to enjoy the car. It’s a car. I don’t care about cars either. I don’t enjoy them as much as use them for their purpose of transportation.
Also literally everyone and every company relies on someone else to provide some services or hardware to enable people to do their job and live. Even this conversation could not even happen without other people making sure internet connections are kept stable. This whole "one person is an island" ideology is not something a lot of people subscribe to.
Some companies don't want to spend time doing software things that are not core to their business, so they pay other companies which sometimes do it better. Most definitely not always, but for some companies it would still be better than some in-house hodgepodge.
And many, many people don’t own a car, not because they can’t afford one but because it makes more sense to rent one or hire a car service when needed. Frankly, my dream scenario is to have enough money to hire a personal driver who is on-demand and at the ready when I need to go out and do things. For plenty of people, that’s anathema. But for me, I don’t like driving (or cooking) and I’d much rather pay someone to do it for me.
The same is true for software. I like tinkering with a lot of my own software, but if I was building my own business, I sure as shit would rather pay for an established company to handle some of the grunt work, rather than hiring dedicated teams that in aggregate would cost a lot of money.
The rise of SaaS and PaaS isn’t just laziness. It’s a recognition that for a lot of people, even developers, we’d rather entrust a lot of the operations and IT work to someone else.
My point is that none of the analogies are arguments for (or against) SaaS, or in this case "brand name" enterprise software.
Whenever any of these comically banal analogies are brought out to sell something, the intention always seems to be to suspend critical thinking.
There exists many legitimate arguments, for many things. They never begin with "why would anyone need a frying pan".
And when putting in the well, why would you pay someone else to dig it for you? Just laziness. Dig it yourself!
And then when digging the well, why would you rent the drill? Just laziness, outsourcing the maintenance and ownership costs to someone else.
And then why bother buying the drill at some market where they're then manufacturing the drill for you? Pure laziness. Just make the drill yourself!
Or maybe, you just go with a trusted service instead of learning how to manufacture digging equipment when all you really want is running water.
Uber is actually cheaper for me than a car + insurance + maintenance. I work from home. I’m going to give my car to my son when we move.
When I need to see my parents 200 miles away, I can hop on a small plane for $300 round trip and just work from there for a week.
Companies these days just want to specialize in making cool menu designs and marketing and waiting on the three star Michelin review to roll in somehow.
That would be like a Hypervisor company paying VMWare to re-distribute ESXi or something.
But a farm company buying VMWare ESXi to run its IT infra instead of using, say, CentOS is exactly like a chef sourcing ingredients instead of growing them herself.
One part of the dynamic, simplified: big companies can show quickest cost savings by laying off employees; this satisfies anxious shareholders during the bumpy times by suggesting that management is wiling to make "hard decisions." Such companies cannot quickly re-tool their technology stacks nor renegotiate 3-year software contracts. B2B companies tend to do very well as long as they are not solely focused on their startup/growth/pre-IPO customers, who get hit hard.
Here are some arguments for post-downturn tech growth: 1981 and the use of the PC + Lotus/WordPerfect/BASIC for efficient local compute 1990 and the surge of the LAN to replace centralized IT 2001 and the swell of SaaS 2008 and the swell of cloud 2022 and the .....
An ERP is basically a database with an interface with business logic built in. Certainly not something you want to do in Excel.
Not just Linux but IT in general. It may have to be about security; my impression is that at least some of the widespread successful attacks in recent years might have been prevented if non IT companies had their own IT department, servers and in house security teams. Relying on an external provider is cheap and comfortable, until the day a single vulnerability screws all its customers data in a single day.
I think you vastly underestimate the cost of on-prem IT, especially for small and medium sized businesses. The "floor" cost of your own IT is many, many orders of magnitude higher than a subscription to something like Office 365 for something like 50 people.
"Normal" people don't use cloud services because they're lazy, they do it because the economics of rolling your own IT does not make sense, and hasn't since the days when it stopped being your only option.
But, even thinking of health care systems is thinking of big businesses. How does a road construction company justify an IT department or security team? Think of businesses like that. They have have an IT guy who manages services they use and is an expert at using those.
Most businesses are small to medium in size and non-tech.
Small businesses with in house It get hacked waaaaasy more.
KVM performance is orders of magnitude better than VMWare and handles migrations snapshots imports and exports without additional byzantine license agreements or mandatory minimums for hardware support on network switches and servers. Cockpit makes it dead simple to run.
Oracle performance is so awful the license terms do not allow you to release performance benchmarks or comparative analysis against other databases. it also has all the same heavy lifting you need to focus on for things like galera clusters or postgres, so theres no clear win unless you like paying Larry for the privilege of slow transactionals on a hyperconverged iron beast, or youre too lazy to figure out ODBC.
And Symantec so openly hates their customers they now bundle a cryptominer with their software. before that their incompetence was so blinding Google had to step in and force them to give up their CA business.
"enterprise" software is an absurd proposition for anyone smart enough to realize their business is more than just the end product. to everyone else, these companies are borderline predatory.
I don't except to the degree that the product is more or less suitable for my business needs. And note I said product, which KVM by itself is not. If I'm going to use KVM--and, yes, I likely would rather than VMware unless I otherwise needed VMware for some reason--I'd be buying it as part of a supported commercial Linux distribution.
The money is of course important, but these providers are smart. They take only what they can and not more. Which is still big money. And while Oracle & co. would never make it into any company I can make a decision for, I don't think that keeps them awake at night - there are enough (big!) fish in the ocean.
We meet on Wednesdays; the coffee and cookies are free.
Do you have any sources for this? I worked at a company developing NFV appliances, we always had much higher network throughput on VMware than we did using KVM without using some type of convoluted vswitch alternative or PCI passthrough.
VMware isn't just a hypervisor, it's an entire ecosystem of VM management and orchestration. You can tie it into AD, delegate different permissions and roles to users/groups, manage upgrades, interact with PowerShell and other APIs, it has integration into Dell and Cisco solutions, all sorts of additional features you won't find running CentOS and KVM without adding more 3rd party software on top and cobbling it together.
This is it, really. For big companies this kind of stuff is important.
And "cobbling it together" is very much understating the effort involved to keep it running: eventually you'll upgrade one of the components and it will break something, because you didn't read the release notes of an upstream dependency that mentioned a breaking change that affects your particular setup.
Having the vendor (vmware) provide this as a delivered, tested, supported solution is so much easier.
Similarly, I suspect it would be significantly simpler to find IT firms and/or hire individuals with VMware knowledge than it would be to find the equivalent on KVM + Cockpit + the dozen other components you need.
Note: I'm not saying this is right or the way things should be, but simply pointing out the "enterprise" perspective. Boring technology is safe.
I have in depth knowledge of kvm and the issue isn’t kvm it’s everything else.
Start comparing VMWare against Proxmox, which is an out of the box solution anyone can use and includes every single feature of ESX and many vsphere features youd easily lose your shirt for. https://www.proxmox.com/en/
heres an independent performance test. KVM is easily faster than ESX.
“No one ever got fired for buying IBM” applies equally to big enterprise SaaS providers and companies like Oracle, Microsoft, VMWare, Salesforce, etc.
It is also deployed in many many enterprise and government settings.
Also, notice when you go to VMWare’s home page you see “referenceable clients” - ie well known companies that use the software? This is “Enterprise Marketing 101”.
Besides, I can throw a stick and find someone who knows VMWare. As a (hypothetical) CTO of a non tech company, I don’t want “support”, I might want an MSP to do it for me.
How vmware is a safe choice is beyond me, their proprietary so one is mostly locked in and at their bidding, their standard support reply is to install from scratch as otherwise the won't look at the setup. If you don't look at the technology POV (in addition to others) the choices probably won't be the best.
Notice how on Proxmox.com there's a testimonial page that links to enterprise and gov't customers, how's that different "Enterprise Marketing"?
Proxmox also has reseller partner all over the world: https://proxmox.com/en/partners/reseller
So we got enterprise support, world wide enterprise reseller network, enterprise features available for all, with or without support, not sure how much more enterprise vibes one can get..
vmware isn't really moving forwards since a while, the acquisition by broadcom won't help that either IMO, but sure they're currently still able to pay "bribe" some CTOs or sales people lunch, so they got that edge going for them, won't be enough to stay the #1 in the long run though.
The average corporation is dependent on 110 SaaS offerings.
https://www.statista.com/statistics/1233538/average-number-s...
No matter what it is a pain to migrate. Corps aren’t being bribed. They are making the safe choice.
For instance even if some unknown cloud provider is “better”, if something goes wrong, the CTO is going to be questioned. If Azure goes down and it’s already an MS shop, people aren’t going to question whether it was a bad decision.
Expensive as hell though.
There are plenty of features Proxmox doesn't have that VMware does have. I've ran into a few of them
1. No ability to pin vCPUs to physical CPUs from within Proxmox. You have to drop to bash and set affinity for each vCPU's PID by hand if you want that.
2. You can't provision a VM with more vCPUs than physical CPUs. For example if I have a host with 8 cores, the max vCPU I can allocate to a VM is 8. And yes, I did have a use case for this.
3. You can't configure networked serial ports from within Proxmox. You have to drop to bash and edit the vm configuration file by hand if you want that.
4. Lack of serial port concentrator, which means you can't really use networked serial ports reliably when migrating VMs across hosts in a Proxmox cluster. In the NFV world this can be pretty important.
5. You can't manage multiple Proxmox VM hosts from a single UI unless they're clustered, which in many cases isn't practical to do. vSphere will let you manage multiple independent hosts from a single pane of glass.
6. (at least historically) lack of RSS/multiqueue in virtio networking. vmxnet3 on VMware supports this and allows you to scale better. But I will admit it's been several years since I've had a look at this area.
Again I'm not a VMware cheerleader. I'm sure I could generate a list like this for what Proxmox has that VMware lacks. But it's incorrect to state that it includes every single feature of ESX.
I've got several clients that pay over 500k/year to a SaaS ERP vendor. The ERP system is the very definition of enterprise software.
What would you suggest I advise my clients when it comes to their finance, inventory, order management, logistics etc?
I guess they could piece together various SaaS solutions to create some sort of composable microservice based system to meet their needs but that's a massive engineering overhead when they can just get all of the functionality they need on one big fat enterprise ERP system.
The cosmetics manufacturer I work with does not care about what the technologists (people like you and me) care about. They just want to run their manufacturing and wholesale operation and if Oracle are offering one system that does it all, why wouldn't they take that deal?
The health care organizations overwhelmingly use third party EHR/EMR systems and schools use third party companies for enrollment (Blackboard).
It doesn’t make sense to bring any those in house.
Bingo! If you are a manufacturing company, you care about two things:
1) does this software do the job I need?
2) does this software cost less than the value I get out of using it?
If the answer is "yes and yes" then -- congratulations! you made a sale!
Sure, if some other company comes in and says "hey, I can do the same job, but cheaper," the customer would listen. But so long as the software works good enough, then the externalities of support contracts, billing, "enterprisey-stuff" may matter more than features or performance.
Remember, not only are most non-software companies better at their business than they are at software, they're also not hiring from the same tech talent pool, they're not paying their tech people as much, they're not letting the tech run their business.
In this case, they have no need for VMWare or Oracle.
Actually, the parent comment is on to something. "Brand name" enterprise software is a buoy that certain types of careerists handcuff themselves to, which allows them to float through their careers fairly unchallenged.
At one point in time, IBM had this market position. Then Microsoft, Oracle, and now Amazon and Google.
Or maybe, just maybe, really big corporations that make software solutions are often really big because their solutions are, if not feature-wise the absolute best, by far the most stable and reliable?
If one were to buy your argument, that would be like saying that businesses buying the Google Apps suite is nonsensical, because the only reason you'd ever use Google is because you're a "careerist" that "only know" Google.
That's obviously not true, Gmail and the rest of the Google suite have been market-leading for many years, because they are good solutions that solve real problems. Presumably the same thing goes for the Microsoft Office suite, and so on.
Just like people did with IBM, Oracle, J2EE, and so on.
If at one point Google Apps is superior to Office 365 makes no difference.
The insight is that the careerists are the effective insider salesmen of enterprise software. Not specs, benchmarks, stats or anything like that.
The idea of product “officers” can be bizarre and counter productive, while that product simultaneously can be one of the best in the market.
I’m not saying that O365 is great, I use it at work and I hate it, but enterprise software has always been like this: it’s not enough to be better, you need to be something like 3x better, but if you are it doesn’t matter if you’re going up against Microsoft. Software is a lot more competitive in that way than a lot of other industries.
If a tech company has achieved both technical and organizational lock-in (or maybe "capture"), specs don't matter.
Specs and benchmarks only matter to geeks.
Otherwise, I'm not sure I would put Google in the mix with Oracle / MS and others. They are firmly non-enterprise, in that it is notoriously difficult to get any support from them, even if you are paying (there are exceptions, yadda yadda...). With Oracle, their products may suck (and they do), but the company knows to answer the phone for their customers, otherwise they won't be able to sell that beefy contract in a few months' time.
> Or maybe, just maybe, really big corporations that make software solutions are often really big because their solutions are, if not feature-wise the absolute best, by far the most stable and reliable?
Oracle? Lol.
When switching is effectively impossible, this is what happens.
Oracle has been buying various application companies with installed bases that are difficult to migrate away from. They are essentially buying customers to milk after putting in their enclosed pasture.
The main change is that MariaDB and PostgreSQL have caught up a lot of ground over the last years, so OracleDB has been losing the edge they have been paid for.
> They are firmly non-enterprise
So you acknowledge you don't really have knowledge of the core enterprise suite Google sells to enterprises, but at the same time you're certain they're non-enterprise?
Google Workspace (the new-ish name for Google Apps) is absolutely enterprise. They definitely have enterprise support contracts, and the few times I've had issues while having a support contract they have been easy to work with.
You hit the nail on the head why enterprises stay away from Google. They are going to lose focus on it as soon as no one internally can justify maintaining it to enhance their careers and show “scope” and “impact”.
https://www.computerworld.com/article/3637079/as-google-move...
> As of March 2020, there were 6 million paid customers, according to Google’s most recent publicly available stats.
There's still supposedly over 2 billion users of these services. Its not a small project with a niche client set where absolutely nobody pays. Reader was popular, but not 2 billion users popular. How many people do you think really used Allo? Its really not in the same space as the many products they've killed.
People on HN don’t grok “the enterprise”.
I mentioned in another post, that even though I work for $BigTech, I work in the cloud enterprise consulting space (not Azure). Dealing with the enterprise is a completely different beast than dealing with other tech companies. Microsoft has almost four decades of experience in the space.
Even on the personal side, I would much rather deal with MS for my one lonely 5 person Office365 subscription than Google support.
Um... perhaps because it is less costly and will lead to better capabilities, more reliability, less complexity than burying it under 18 layers of apis, containers, and virtual machines.
Over the years, I’ve come to suspect that maybe the technology is less of a differentiator in the overall effectiveness of a company’s IT environment than the people running it.
And with IT staffing supply having lagged behind demand (as also evidenced by relatively high salaries in IT vs many other professions), you could easily argue that it was (still is?) historically more difficult to hire good IT people than in many other professions.
And arguably that’s making outsourcing of IT needs relatively more attractive than outsourcing of other business functions where reasonably competent leadership and staffing is easier to come by.
And arguably, capturing many of the benefits of Open Source require more IT competency than using off the shelf mainstream commercial software. If I’m using the same crap that everyone of my competitors is using, none of us win or lose on that. But when using Open Source, I’m more likely to depend on the quality of my IT leadership and staff to outcompete my competitors. And unless a CEO is quite IT literate, s/he probably doesn’t want that additional headache of becoming good at figuring out how to hire good IT leadership.
So I’d hazard a guess, that insourcing and the use of Open Source will become more attractive for many corporations only if/when IT salaries drop more in line with other professions.
https://www.crowdstrike.com/cybersecurity-101/endpoint-secur...
(the VMWare owned product isn't CrowdStrike, that just had the best SEO page with a clear description. VMWare have Carbon Black which is the other dominant player in this market https://www.vmware.com/uk/products/whats-new/carbon-black.ht... )
I work for CrowdStrike just so any potential bias is out there in the open, but do not speak for them.
My role also encompasses many “CIO-like” functions, but my area of expertise is in application development and I look to hire strong leaders in the other areas (or move certain tech domains entirely into other functions — eg our warehouse management system lives within our operations group, and me and my team help in an advisory capacity but don’t “own” the budget or WMS strategy).
I can't imagine it being cheaper to build everything yourself on top of just plain KVM, or even LXC.
You can see people spending millions of dollars a year to keep from having to deal with Linux as evidence that they are stupid. Or you can see it as evidence that Linux really is that bad. It works both ways.
VMware gives you a thing people can be trained on and certified on, a brand you can hire for and screen resumes on, a consistent environment which behaves in a predictable way that you really can turn employees into replacable cogs. Any helpdesk or admin employee can deal with VMware, any MSP, any tech recruiter, and a lot of training companies. You can get backup systems which "support VMware" and storage which integrates with VMware snapshots, and reporting tools which work with VMware.
It's almost not about the tech at all, it's about how do you build companies on shifting sands? You define interfaces for components which can be plugged together. "VMware" is an API or interface that the business can work to; vendors can say "deploy this OVF to VMware", sales can say to the business "this thing we need works with VMware" or to the customers "we can work with your VMware" or "our offering is trustworthy because we use VMware" and the customer recognises the name. HR can say "we need to hire people who know VMware" and that means something fairly specific to the wider world. "runs on Linux" and "people who know Linux" are wildly, wildly, variable and vague things which could mean "ran a website, minimum wage" or "turns SELinux off to make things run" or "was SRE for FAANG" or "did a PHD in AI for tuning networking stacks in HPC applications but doesn't know anything else".
You make software by defining interfaces and components that can be plugged together to make larger systems. Brands are that, for tech. Like you hire someone who "knows React" not someone who "is a programmer" because that's too vague and is as likely to get you someone who worked on a Java CRUD program or someone who worked on a Python log analyser. Like you hire a "service delivery manager" or a "customer account manager" and not "an employee".
Broadly speaking, anyone generalising the world into suits and techies needs a translation layer between them and anyone else.
A friend uses VMWare in an IT role at a medium sized company. They don’t have a software engineering team, just a bunch of users and an IT skeleton crew; so it makes sense for them.
Microsoft has a monopoly on Corporate Domains/OS as it is.
Have you ever considered the other way around? That is, “What changes with age that would make someone over 40 prefer outsourcing a problem to an enterprise?”
It’s possible that this trend is a result of a different perspective.
Something I saw in action building a startup is that the fewer people you have on your team, the more efficient they are.
As you grow, it’s wise to focus your headcount on your core competency and outsource everything else so that you are committing the organizational expense of more employees to all and only what you need as a business.
Of course, enterprise stuff is not all roses. You still have to have someone with product expertise. There’s vendor lock-in. Enterprise software is often oversold. Etc etc.
There is no panacea, just stuff that works better in different situations.
I worked at companies as a software developer from 1996 - 2012 that had to manage their own infrastructure. But today, the only company that I worked for back then that would be managing their own infrastructure today is the one that has mainframes and hardware that handle the backends for lottery systems across the US.
By 2012, there was a slow shift to the cloud.
I first was exposed to how large enterprises worked in 2017. I was hired to lead two green field implementations. But at the last minute they decided to “move to the cloud” neither they nor I knew anything about the cloud. They hired consultants and a Managed Service Provider. Of course the internal IT department was vigilantly defending their turf and the “consultants” were old school Netops folks who only knew how to “lift and shift” and duplicate an on prem infrastructure and all of the red tape to the cloud and of course it was more expensive than just using a colo.
I spent the next six months after the decision was made studying AWS and getting a certification not because I value certifications (I don’t). But it gave me a guided learning path to know what I didn’t know. It did open my eyes to what I wanted to do - work with companies - specifically developers and operations to show them how to actually take advantage of cloud and not just do lift and shifts - ie true “Devops”.
I left that company and went to a small startup for two years where I learned everything I know about “cloud application modernization” and then ended up in Professional Services at AWS.
Until I started working with large enterprises and government organizations from the consulting side, I never appreciated the concerns of large enterprises and how they aren’t in the “tech” business and it does make sense to outsource that knowledge - not to ProServe we don’t do that type of work - to external partners.
As far as VMWare, as silly as it sounds on the surface. Companies actually use VMWare to manage hybrid infrastructure on the cloud and on prem as a “single pane of glass”.
https://aws.amazon.com/vmware/
I personally don’t deal with those implementations. I stick with app dev.
New companies, too. There's a lot to be said for paying someone to make problems go away. Not everyone wants to write their own software or change their own oil.
"Use proxmox" - fucking lol
Yes I know how to do app dev + cloud Net ops, it’s kind of my thing.