This isn’t about the hypervisor. This is about the enterprise sales. A bank isn’t going to run their data center on qemu, regardless of the performance. They are going to go with the company that can sell them support contracts with dedicated engineers to answer the phone.
Big companies are starting to learn that free open source can be supported in house, in some cases it is even cheaper, but in others the flexibility of being in control of when a much needed by only you feature is developed is a big deal. Open source also is never acquired and the price jacked up to beyond the worth but they are forced to pay the price anyway because they can't get off before their license expires.
> Big companies are starting to learn that free open source can be supported in house
Sure but the problem here is can you find and retain the people who can do that work, if you are a big lumbering "un-interesting" company where tech is not your core focus? In most cases the answer is "no".
There's a reason major software vendors provide professional services. If their customers could reliably hire and (most importantly) retain experts on their own, they wouldn't need professional services.
>There's a reason major software vendors provide professional services. If their customers could reliably hire and (most importantly) retain experts on their own, they wouldn't need professional services.
This. As a tech person with decades of professional services experience in the tech space, I couldn't agree more.
Generally, there are three reasons why corporate entities hire professional services (this isn't limited to tech stuff either):
1. There's a specific need for design/implementation expertise where FTEs are either not readily available or are unnecessary once the implementation is complete;
2. The internal group in favor of a particular technology/implementation wants/needs validation from "experts" to get full buy-in from management;
3. Professional services are brought in to "validate" existing project plans/implementation details as a CYA against project failure (e.g., "the consultants told us this was the way to go. They were wrong, not us.") which (both bringing in consultants and project failure) happens a lot more often than you might think.
Those are all valid points. However those are also the counter argument: eventually the cost of all those services is more than the cost to do it in house, or find a competitor.
Yeah but again, it's most often not about the money...
I work for a large enterprise and we pay or retain vendors to do things for us all the time, at a much higher cost than it would be to do it in house, because we cannot find/hire/retain the expertise to do it ourselves, and at some point we need to get the work done so that other more mission-critical work can happen to drive the business forward.
Plus, there is value in paying someone to make the find/hire/retain part their problem instead of ours.
We buy the reliability of getting the work done in a timely manner, on top of the work itself.
I work for a small company and they seem to have figured this out 20 years ago. I've been tasked with supporting the network and I've got to say that all the fancy (or not) proprietary firewalls all seem like a pick your own poison of cost/support issues, and they all go end of life eventually but the network needs to keep running. We also use PFSense in a few places and I think the feeling among the operations team (2.5 of us) is that we'd like to just use that in future.
I use PFSense and Snort at home and it's perfectly fine for the price I paid for it, but enterprise grade it ain't. I wouldn't use it to protect a network worth more than my beer budget.
I don't know what business you're in and maybe after a nightly tape backup there's nothing important for you to lose. Frankly that's most small businesses. But if you've got real trade secrets get a proper IDS/IPS before you regret it.
Sure, the networking thing is a part time job for all of us, and there's even been some talk of bringing in professional services to look over it a bit. But my question is, what would we get? The whole IDS/IPS/WAF industry seems pretty much dripping in snakeoil. It seems like Cisco was once the king, but has lost a lot of respect. Palo Alto seems to be the current fashion but unless you have a serious alcohol problem, it's going to exceed your beer budget. I worked with Imperva and Alert Logic back in the day and no one ever seemed too impressed with them. Then there's SonicWall which seems reviled and Fortigate which seems like the reasonable budget option. I'm sure there's tons others I'm not even aware of.
> They are going to go with the company that can sell them support contracts with dedicated engineers to answer the phone.
There's no reason the company selling them support couldn't be providing support for QEMU btw. If it met their needs otherwise I'm sure they'd consider it.