They are not the only ones. The company I work for uses a LOT of VMware. Mind you, VMWare has ALWAYS been expensive. To the point that in a lot of cases, you can end up spending more on VMWare than the hardware that runs it. I long ago crunched the numbers and came to the conclusion that unless your workload really does require the kind of exceptional storage, networking, and HA integration that VMware (claims to) provide, you are better off using an open source solution for your VM infra and hiring a couple of extra engineers to manage it.
When it was time for our contract to renew a few months back, our licensing and support quote was 3x the previous year and Broadcom would not budge, even a little. They said take it or leave it. Well, we left it.
There is a big internal effort now to get us off of VMware and onto Kubernetes and OpenShift. Our whole fleet of VMware is still running but we're on borrowed time as we're on our own if any major technical issue comes up.
I didn't see the numbers but at our company the savings of moving virt to OpenShift from VMWare were (allegedly) significant. Plus there was something about not having to renew subs for RHEL on the VMs that would be a cost savings as well. So I guess it depends.
Not necessarily. Red Hat negotiates prices with customers. If your company is paying as much for OpenShift as they would/were for VMware, they need to hire better negotiators.
How good is VMware anyway? I’ve only ever “used” VMware in the consumer use case… I think it was a demo version (maybe there was a free student version over a decade ago?). Anyway, it seemed much worse than VirtualBox (the options were less intuitive and the GUI was ugly) but then I don’t think I the user they were targeting (student playing around in Linux).
> the options were less intuitive and the GUI was ugly
Oh I see you tried our enterprise software. Welcome!
VMware is bananas. One console runs thousands of VMs across hundreds of physicals in dozens of data centres. VMs can, and do, move around at will thanks to vMotion. Need to upgrade this physical host? Just vMotion its guests somewhere else.
Oh and it’ll replicate all of this for you, live. So if one half of your data hall goes down, it doesn’t matter.* Your users don’t even notice.
And many, many other features that Virtualbox can’t touch.
(*Well, someone like me is having a bad day, but you know.)
My $dayjob is freaking out about the new cost of VMware.
I suspect Broadcom's plan is to figure out just how much they can bear and cut a deal along those lines.
That would totally work -- we're neck deep in vmware, and there's no easy way out -- except that there's a kind of network effect at play.
Support is huge for us. In the beginning, we developed expertise in vmware (particularly in regard to our products), and guided our customers to it as well. Customers were happy to follow our recommended/supported option, and we were happy supporting it. (The support ain't cheap!)
But our customers, by-and-large, are smaller and somewhat more nimble, meaning many will be unable or unwilling to do a deal with broadcom, and will insist on switching away. I suspect new customers will lead the way... there's no way we're going to walk away from a deal because the customer refuses to pay for vmware, and we probably can't afford to eat that cost for them. So we're going to start gaining expertise and supporting other vm platforms. I think we'll follow our original pattern: settle on one, gain expertise, start using it in support, then test and dev, and steer our customers to it. At some point we'll realize we know how to get rid of the rest of our internal use of vmware, AND realize we can sell remaining customers support to switch away from vmware themselves. Once that happens, I think it becomes a ball rolling down hill, and things will accelerate quickly.
There are a couple reasons this isn't like Oracle, BTW. For one, with Oracle we're 10,000 ft below the surface with Oracle-provided air tanks. Second (I guess this is the main thing), Oracle is our problem, not our customers'.
I have to wonder if stuff like this will, over time, force companies to be extremely wary of any tech that eventually falls into the "extremely difficult to remove" category, and have companies instead requiring open source solutions and then separate contracts with support companies for those solutions.
It's just further re-emphasizing that it doesn't matter how good your vendor is now, they will probably eventually get acquired by a company attempting to squeeze every last penny out of you. And their cost-benefit calculation is no longer based on how much unique value the software is providing, but it's instead about how much of a PITA it is to migrate off said software.
Dating back 10-15 years, that premise was exactly what drove the enterprise shift away from a Microsoft stack (.Net + SQL Server + Windows Server) to a LAMP stack (where the L-A-M-P were each substitutable by various OSS options). It required a lot of reskilling in the enterprise, but after everyone was up to speed on the fundamentals of the new architectures and comfortable programming in Java/Python/Javascript/PHP it made learning all the newer things to come even easier. It also set the stage for companies like Canonical and Redhat to create viable commercial businesses wrapping FOSS software in enterprise support.
Where we are now is on the cusp of starting another of these lock-in rejection cycles, but where the lock-in isn't at the OS or devtools layers, but at the data/analytics layer and with IaaS or PaaS having become unreasonably expensive alternatives to on-prem data centers [for many enterprises].
It'll be interesting to see how things evolve for Snowflake & Databricks (as well as Pega, C3AI, and similar), and whether CIOs start placing bets on their own team creating their own solutions using FOSS tooling on-prem, leaving public cloud as the domain of things like ERP & whatever SaaS business software they license.
You just described vendor lock-in. It's never fun, but it's usually a known factor for any competent decision-makers. Sometimes "good for now" is good enough...
Umm. The contingency plan for any big corp is to continually plan to migrate off a legacy system. It takes forever for big corp, if they don’t have a contingency plan now they will never have it. He wants to drop a bunch of run books on people and tell them go do that. He only wants configuration and does not want any development.
They are looking to extract the maximum amount of money possible. I'd argue that Broadcom could extract more money with smaller uplifts but I think they are also looking to consolidate their customer base. Some of these crazy numbers may be doing just that - saying that these people aren't wanted as customers anymore.
10x price increase? Sounds like a classic Computer Associates move. I think when Broadcom bought CA they also acquired this aspect of CA’s culture, and are now applying it to products which never had anything to do with CA.
To think that the migration away from VMWare would cost $40-50m but have a "very quick payback" presumably means that AT&T is a gigantic customer paying a ton of money, so if they're not in the top 10% contracts worth keeping then who is?
Looks similar to what was Boeing and McDonnell Douglas. They even said McDonnell bought Boeing with Boeing's money. Looks like CA bought Broadcom with Broadcom's money
If you see an enterprise software product starting to decline, as VMWare has in the cloud era, then sometimes the "best" (but evil) thing to do is just milk your customers as hard as you can. Just squeeze every penny out of people who are stuck with you. Any migration of a large enterprise will take several years so you can just get all that money in one big go and then shoot the product in the head.
If you see a tenant starting to complain, like that one who’s been in your building for years, then sometimes the "best" (but evil) thing to do is just milk 'em for all they’re worth. You jack up the rent—ten times what they’re paying now. They ain't got no choice. Ain’t no place for them to go, so you just squeeze every penny outta ‘em. It’ll take ‘em years to find a new spot, so you got plenty of time to take all that cash in one big lump. And once you’ve wrung ‘em dry, you just tell 'em, 'Thanks for your business,' and throw 'em out. Nothing personal, kid. Just how the game’s played.
In the mid-2000s I worked for a place where we were paying exorbitant annual license fees for this CA eTrust LDAP we hardly used. We mainly used Novell eDir and Sun JES LDAP, we were so happy when we could migrate the eTrust to Sun. Later on, Sun JES got replaced by Oracle Internet Directory, and Novell eDir got replaced by AD.
Say what you like about Sun or Oracle or Novell, their pricing was much more reasonable than CA’s. Plus we were a public university, and CA didn’t seem to believe in education discounts, whereas Oracle gave us a standard education discount of over 90% off list price.
CA was famously the place where mainframe software went to die. When I worked for Oracle, I had some very limited exposure to CA TopSecret and ACF2, which are mainframe security products (RACF competitors) that CA bought, which an Oracle product I was working on integrated with. No idea what the licensing was but I’m sure it wasn’t cheap.
What you're describing is more like private equity, not a hedge fund. Hedge funds are generally passive investors and don't usually drive product pricing in their portfolio companies.
Can someone explain what VMWare does so well that allows them to do this for someone totally uninformed? No where I've worked at has used VMWare, but everywhere I've worked used containers or VM's of some sort.
Even in the early 2015's kubernetes was mature enough to run production workloads. Any issues that did come up were with our own application code, never the orchestrator.
I never found myself thinking that it would be nice to have a close sourced expensive option to reach for.
What area does VMWare excel so much to justify this pricing power?
VMWare is far easier solution for normal enterprises than K8s. K8s more suited for having many small VMs that can be quickly deleted and recreated i.e modern microservice architecture. vSphere & friends is more targeted for running very large database, oriented application that need high uptime and are very long lived. VMWare can live migrate a running OS between physical hosts so that you can have continuous uptime. VMWare works with any OS so it's especially used by any Microsoft based orgs which the majority of hospitals, schools, government offices are.
If you are deploying enterprise apps from the 1990-2000s you use vSphere, if you are building your own SaaS product then you use K8s.
Kubernetes demands a cloud-native approach - VMs are deployed from container images, nothing has persistent storage by default, network interfaces must be declared, etc. VMWare tolerates more conservative approaches like hoisting existing physical machines into VMs while leaving everything running on them unchanged. It's less sophisticated on a technical level, but it's also an easier "sell" for companies with an established technology stack that they may not want to change right away. It's also potentially a better fit for organizations which need to run Windows infrastructure (e.g. AD servers, mail servers, etc) which can't be hosted in k8s.
> What area does VMWare excel so much to justify this pricing power?
My experience is only with VMWare's desktop-virtualization tools, but hands-down they have the best integration features and services, especially for... uh... "retro" small-business computing needs (in my case, it was the only way I could get a VM running Windows Server 2003 to work - which I needed in-order to be able to run a Progress-based CRM).
I find it odd that Microsoft's own virtualization/Hyper-V stuff is useless if you're wanting to run older versions of Windows, especially XP/2000/2003 (as Hyper-V was a post-Vista/WS2008 thing); it's not just the lack of drivers, but the lack of absolutely-essential integration features like USB port forwarding and "real" GPU emulation (because Hyper-V's "Enhanced Session mode" doesn't actually show you the local-console desktop: it's all just using a special mode of RDP).
If you actually literally meant "or VMs" in your statement, I'm curious what sort they were that were not VMWare?
VMWare absolutely owns this market in the enterprise. It's been reliable, it's well-supported, there's a huge ecosystem of vendors and integration partners around it, and it's been the no-brainer choice for virtualization for CIOs since about 2010 (or earlier!).
Yeah so we used Vagrant way way back in the day to have an easy way to distribute dev environments. Definitely not an expert here as to all the alternatives, but for what we had, the devex felt fine.
I've been a user and admin (not architect or SME) of VMware for over a decade and what strikes me is the reliability.
With that said though, I think a lot of customers using VMware don't really need all the features of VMware. They could get away with something simpler and cheaper.
Still don’t get what the point is. You can temporarily squeeze the locked in customer for good quarterly numbers but in long run youre just fucking up your business permanently
What is "long run"?
Oracle has been doing this for decades.
For some companies, the migration from oracle to anything else would cost too much and hard to justify in some places, especially government.
It's the same for VMware, It will takes many years for some government agencies to replace it with something else.
I'm guessing Broadcom is going the same way.
They can't compete against cloud migrations or open source alternatives.
So as you said, they're squeezing locked in customers, the fastest they can until they've all migrated... Or until some decided that keeping it at that price was still a better ROI than migrating.
I don't think a lot of people will start a new VMware data center in the years to come.
That was Broadcom's plan all along, they weren't buying VMware to compete, they were buying it to wring more than $61Bn from the customers who are stuck.
This[1] from 2022 says "Broadcom's stated strategy is very simple: focus on 600 customers who will struggle to change suppliers, reap vastly lower sales and marketing costs by focusing on that small pool, and trim R&D by not thinking about the needs of other customers – who can be let go if necessary without much harm to the bottom line. The Register offers that summary based on Broadcom's own words, as uttered at a November 2021 Investor Day."
and "Krause said Broadcom is content to have those 100,000 customers "trail" over time."
> Or until some decided that keeping it at that price was still a better ROI than migrating.
And then at the next renewal, when the price goes up 3x, the cost of a one-time migration will look cheap - to the new executive that's in charge at that time.
Execs meet their target to get their outsized stock options. The Board gets theirs too. The short term investors got theirs too. It is irrelevant that the company is a burning wreck of its former self and customers and employees are saying WTF. This is the reality of Shareholder Primacy.
It really does seem as though Broadcom is entirely shifting VMware's focus the the top 5 or 10 percent of customers who probably make up the vast majority of the actual profits. The message they've delivered time and time again to businesses outside that group is pretty simple "go away" price.
It seems like it, making VMware Fusion and Workstation free seems to fit with that strategy.
A bit of nostalgia: my first VMware product was VMware Express for Linux. It was a stripped down version of Workstation (probably 2.0?) that could only run Windows 95/98:
Readit News
When it was time for our contract to renew a few months back, our licensing and support quote was 3x the previous year and Broadcom would not budge, even a little. They said take it or leave it. Well, we left it.
There is a big internal effort now to get us off of VMware and onto Kubernetes and OpenShift. Our whole fleet of VMware is still running but we're on borrowed time as we're on our own if any major technical issue comes up.
Oh I see you tried our enterprise software. Welcome!
VMware is bananas. One console runs thousands of VMs across hundreds of physicals in dozens of data centres. VMs can, and do, move around at will thanks to vMotion. Need to upgrade this physical host? Just vMotion its guests somewhere else.
Oh and it’ll replicate all of this for you, live. So if one half of your data hall goes down, it doesn’t matter.* Your users don’t even notice.
And many, many other features that Virtualbox can’t touch.
(*Well, someone like me is having a bad day, but you know.)
I suspect Broadcom's plan is to figure out just how much they can bear and cut a deal along those lines.
That would totally work -- we're neck deep in vmware, and there's no easy way out -- except that there's a kind of network effect at play.
Support is huge for us. In the beginning, we developed expertise in vmware (particularly in regard to our products), and guided our customers to it as well. Customers were happy to follow our recommended/supported option, and we were happy supporting it. (The support ain't cheap!)
But our customers, by-and-large, are smaller and somewhat more nimble, meaning many will be unable or unwilling to do a deal with broadcom, and will insist on switching away. I suspect new customers will lead the way... there's no way we're going to walk away from a deal because the customer refuses to pay for vmware, and we probably can't afford to eat that cost for them. So we're going to start gaining expertise and supporting other vm platforms. I think we'll follow our original pattern: settle on one, gain expertise, start using it in support, then test and dev, and steer our customers to it. At some point we'll realize we know how to get rid of the rest of our internal use of vmware, AND realize we can sell remaining customers support to switch away from vmware themselves. Once that happens, I think it becomes a ball rolling down hill, and things will accelerate quickly.
There are a couple reasons this isn't like Oracle, BTW. For one, with Oracle we're 10,000 ft below the surface with Oracle-provided air tanks. Second (I guess this is the main thing), Oracle is our problem, not our customers'.
It's just further re-emphasizing that it doesn't matter how good your vendor is now, they will probably eventually get acquired by a company attempting to squeeze every last penny out of you. And their cost-benefit calculation is no longer based on how much unique value the software is providing, but it's instead about how much of a PITA it is to migrate off said software.
Where we are now is on the cusp of starting another of these lock-in rejection cycles, but where the lock-in isn't at the OS or devtools layers, but at the data/analytics layer and with IaaS or PaaS having become unreasonably expensive alternatives to on-prem data centers [for many enterprises].
It'll be interesting to see how things evolve for Snowflake & Databricks (as well as Pega, C3AI, and similar), and whether CIOs start placing bets on their own team creating their own solutions using FOSS tooling on-prem, leaving public cloud as the domain of things like ERP & whatever SaaS business software they license.
They are looking to extract the maximum amount of money possible. I'd argue that Broadcom could extract more money with smaller uplifts but I think they are also looking to consolidate their customer base. Some of these crazy numbers may be doing just that - saying that these people aren't wanted as customers anymore.
For those that do not know the reference, "Fork Yeah! The Rise and Development of illumos" by Bryan Cantrill at LISA11:
* https://www.youtube.com/watch?v=-zRN7XLCRhc&t=38m24s
And the lead up is entertaining too:
* https://www.youtube.com/watch?v=-zRN7XLCRhc&t=34m7s
That said, even AT&T is jumping ship: https://arstechnica.com/information-technology/2024/10/broad...
To think that the migration away from VMWare would cost $40-50m but have a "very quick payback" presumably means that AT&T is a gigantic customer paying a ton of money, so if they're not in the top 10% contracts worth keeping then who is?
https://www.ciodive.com/news/broadcom-att-vmware-settlement-...
Say what you like about Sun or Oracle or Novell, their pricing was much more reasonable than CA’s. Plus we were a public university, and CA didn’t seem to believe in education discounts, whereas Oracle gave us a standard education discount of over 90% off list price.
CA was famously the place where mainframe software went to die. When I worked for Oracle, I had some very limited exposure to CA TopSecret and ACF2, which are mainframe security products (RACF competitors) that CA bought, which an Oracle product I was working on integrated with. No idea what the licensing was but I’m sure it wasn’t cheap.
Even in the early 2015's kubernetes was mature enough to run production workloads. Any issues that did come up were with our own application code, never the orchestrator.
I never found myself thinking that it would be nice to have a close sourced expensive option to reach for.
What area does VMWare excel so much to justify this pricing power?
If you are deploying enterprise apps from the 1990-2000s you use vSphere, if you are building your own SaaS product then you use K8s.
My experience is only with VMWare's desktop-virtualization tools, but hands-down they have the best integration features and services, especially for... uh... "retro" small-business computing needs (in my case, it was the only way I could get a VM running Windows Server 2003 to work - which I needed in-order to be able to run a Progress-based CRM).
I find it odd that Microsoft's own virtualization/Hyper-V stuff is useless if you're wanting to run older versions of Windows, especially XP/2000/2003 (as Hyper-V was a post-Vista/WS2008 thing); it's not just the lack of drivers, but the lack of absolutely-essential integration features like USB port forwarding and "real" GPU emulation (because Hyper-V's "Enhanced Session mode" doesn't actually show you the local-console desktop: it's all just using a special mode of RDP).
VMWare absolutely owns this market in the enterprise. It's been reliable, it's well-supported, there's a huge ecosystem of vendors and integration partners around it, and it's been the no-brainer choice for virtualization for CIOs since about 2010 (or earlier!).
With that said though, I think a lot of customers using VMware don't really need all the features of VMware. They could get away with something simpler and cheaper.
For some companies, the migration from oracle to anything else would cost too much and hard to justify in some places, especially government. It's the same for VMware, It will takes many years for some government agencies to replace it with something else.
I'm guessing Broadcom is going the same way. They can't compete against cloud migrations or open source alternatives.
So as you said, they're squeezing locked in customers, the fastest they can until they've all migrated... Or until some decided that keeping it at that price was still a better ROI than migrating.
I don't think a lot of people will start a new VMware data center in the years to come.
This[1] from 2022 says "Broadcom's stated strategy is very simple: focus on 600 customers who will struggle to change suppliers, reap vastly lower sales and marketing costs by focusing on that small pool, and trim R&D by not thinking about the needs of other customers – who can be let go if necessary without much harm to the bottom line. The Register offers that summary based on Broadcom's own words, as uttered at a November 2021 Investor Day."
and "Krause said Broadcom is content to have those 100,000 customers "trail" over time."
[1] https://www.theregister.com/2022/05/30/broadcom_strategy_vmw...
And then at the next renewal, when the price goes up 3x, the cost of a one-time migration will look cheap - to the new executive that's in charge at that time.
https://en.m.wikipedia.org/wiki/Shareholder_primacy
https://en.m.wikipedia.org/wiki/Friedman_doctrine
https://www.sciencedirect.com/science/article/pii/0304405X76...
Then VMware dies because it cannot decrease prices anymore due to lack of volume.
A bit of nostalgia: my first VMware product was VMware Express for Linux. It was a stripped down version of Workstation (probably 2.0?) that could only run Windows 95/98:
https://web.archive.org/web/20010124081300/http://www.vmware...
Does anyone remember Win4Lin 9x (based on SCO Merge)?