We (Red Hat) are doing great business moving customers off VMware. 20,000 VMs is not an unusual amount for our larger customers. I did a 5 min lightning talk in summer about this: https://pretalx.com/devconf-cz-2024/talk/SN93LG/
It is funny because my former company was trying to move out of vmware even before the broadcome aquisition and the local sales rep pretty much told us RHEV was a dead project but at the time openshift virt wasn't ready yet.
In the end I left before broadcom bought vmware but the status at the time was "let's wait a bit more".
I don't agree with the deprecation of RHV since it's very closely modelled on VMware vCenter in terms of the concepts and user interface. So it's ideal for ex-VMware admins. But here we are. Containers & Kubernetes are more buzzword-compliant than virtual machines.
Basically, RHEV as a VMware takeout project never gained a lot of traction. With the general industry shift towards containers and Kubernetes plus the Broadcom acquisition, kubevirt in combination with a much more mature OpenShift is a lot more interesting.
Looks like that tool typically targets KVM as a source. My understanding is that Red Hat's own virtualization platform is transitioning from RHEV to OpenShift. If that is true, are you actively moving customers from vmware to OpenShift via Kubevirt?
Upstream, virt-v2v supports conversions from VMware to either oVirt (RHV) or KubeVirt (OSV), and we don't plan to drop the oVirt support any time soon.
However Red Hat now only supports conversions to OSV, since RHV was deprecated (sadly).
The biggest problem is oVirt itself has not proven to be a very sustainable open source project. If oVirt dies, we'll likely remove support in v2v. (There's a great start up opportunity here, for a dull but money-making company that productizes oVirt again.)
To move VMs from RHV to OSV you can just copy the disk image since they already should have virtio drivers, qemu guest agent, and be able to boot on any qemu/KVM-based platform. I believe there's some automation for that, but it doesn't involve virt-v2v.
Broadcoms policy seems to be to aim for the larger customers where a 10x price rise for a suite of products isn't an issue. Charging 10x the price to 10% of your customers isn't necessarily a bad strategy for a business.
> Steve McDowell, chief analyst at NAND research, told The Register that VMware by Broadcom is “laser focused on high-revenue, high-margin business” and has priced its wares “just below the pain threshold for customers they care about.”
There is a famous story about business in the UK when some sales manager at Coca Cola raised the price by a small amount for JD Wetherspoons, one of the largest chains of cheap pubs across the country. The manager must have felt very smug about the extra revenue they'd bring in, from a small change that the client could easily absorb and could hardly moan about. What else could they do, move to Pepsi?
They did exactly that, and the sales manager was fired over losing one of the biggest clients and the loss of revenue that came with it.
One needs to be very, very careful about both the pain threshold and the "** you, I'm moving my business elsewhere" threshold.
But if Broadcom increases price 10-fold and loses half the customers by revenue over the next couple of years, that reflects in their bonus and they can then move on.
The customers might have a plan to move off vmware by 2027, but continue to pay the money until then. Come the 2027 renewal the execs that made the massive bonuses for increasing revenue have left and then when the company says they aren't renewing it's the next generation dealing with the problems, and then they just reach for the three-envelopes.
This really is on the companies that choose a single supplier. If you had two providers, then shifting doesn't take a 3 year program, it's just a policy change to prefer Provider B rather than Provider A.
And most importantly some clients will not communicate that they are unhappy and just move on quietly. Especially for something like virtualization which is essentially a commodity.
I would guess the plan for the larger companies is to sell a whole suite of products around the VMware: management, ops, auditing etc, plus take very good care of their niche requirements so moving becomes a much more expensive proposition
But does it not make you more vulnerable to competitors? Losing one of your few big customers will hurt regardless, but it likely hurts less if you have a pool of SMBs to act as a buffer.
In terms of competition it's really the same. Whether you fight with a team of 1 for 5 businesses or a team of 10 for 1 business that is 50x bigger, on average as long as you have equal competition in each segment you walk away with the same. But you get very different customers, for better or worse.
A segment of a customer base is hardly a buffer, if you underserve them they will leave, and you will have wasted effort making an inclusive sales and support process to them.
i never understood, even 15 years ago, why people still use vmware when so many open source (i.e. virtualbox) systems existed at the time. The fact that Broadcom buys it out and then 10x the price on people is just even more reason to move to open source. As noted in the article, the open source solutions are 2x faster anyway, so they customers can realize a 2x increase in slices per server.
The ones staying with VMware, permanently or near term, are likely using products for which there are not good replacements. The open source alternatives can do what ESXi does sure, but its progressively more difficult to replicate VSAN (including stretched clusters) and NSX-T (with distributed firewalling) without bringing in multiple vendors that will all point at each other when there is a problem.
There are some open source options that come close to what vmware's offerings could do, but virtualbox for sure ain't that. It has an extremely limited feature set compared to other offerings in this space. For example, it is mainly intended to manage VMs on a single host. Most organizations looking at commercial offerings have multiple hosts (or racks, or even datacenters) they want to manage. Apples are much more similar to oranges than vmware is to virtualbox.
Hell, AT&T just settled with Broadcom. Five bucks says they stay right where they are. Hock Tan knows what he’s doing to force the big players to pay and extort them to the nines.
A less charitable (but I'd argue, likely more correct on balance of probabilities) is they're aiming for the larger customers who are so entrenched in their use of the VMWare product stack that they can't feasibly migrate without taking a huge one-time hit to their balance sheets.
The desktop is mostly a pretty uninteresting space in general. Virtual desktops never really took off (which VMware was into at one point) and desktop virtualization is generally a commodity for the relatively few people who use it.
not smart if you make it hard to use and learn and the next generation of technologists gain experience and familiarity with other solutions and improve them to be superior products.
Bold of you to assume Broadcom cares about long term sustainability.
From what I have seen they are just a PE firm pretending to be a tech company. They will bleed the customers who can’t move fast enough. Burn it to the ground to get every last dollar.
Unless they have a monopoly on the tech, that's not how the free market works. Raising your prices x10 without having a government mandated monopoly, is a great incentive to create competitors.
From Broadcom's perspective, that seems to make even less sense than a simple price hike that lets them shed unprofitable customers. What's the point of doing that? Some kind of sales force optimization?
A sale isn't free. It requires time, headcount, and engineering+PM resources.
At Broadcom size, you want to maximize sales to F1000 customers (the kind that can and will shell 7-8 figures in TCV) as that gives you the most bang for your buck from a SalesOps perspective.
A major issue with VMWare was their sales team would chase logos at the expense of an optimal sale, so their account teams would give extreme discounts (70-80%) at barely above the margin in order to close deals.
Everyone now has to ask the question, should i continue paying for product that does XYZ from VMWare's competitor when i'm already getting a product that does XYZ as part of the VMWare licence.
Chances are if you're a vmware estate, those products are going to integrate better and then you only have a supplier, support portal, customer rep, etc..
Broadcom loose customers who arn't really invested in to VMWare and those who stay become even more invested in to them.
I work at an IT outsourcing company that also has some VMWare farms (though not sure if we reach 20k VMs).
What I picked up is that before the acquisition, price was mostly determined by CPUs on the hypervisor, now it's mostly by RAM.
Which means we now have to optimize hypervisor hardware to a different licensing model.
I don't think we've seen 10x price hikes, but it could possible if a provider has optimized hard for a metric that is now less relevant, and/or has difference licensing deals.
Caveat: I've picked all that up second-hand through the grapevine, I'm not working with our virtualization farms directly. So not 100% on top of what's going on.
It's the Oracle pricing model. Turn the customer upside down and keep shaking until you've got all their money. They will dress this up in different ways to different customers, but that's the model.
Having been in the field for some time and having some contacts, I must say: 100% of the companies where I get some insights are currently either already building alternatives or reviewing their options with a strong intention to leave VMware behind. And in many cases this is quite a huge task, but certainly doable with the required momentum.
I'm in small Devops team and people are pissed off. Thankfully Proxmox works just as well for us (some would say better). Even a large-scale migration seems scriptable to me.
Proxmox looks great, but unfortunately it doesn’t come with encryption oob and manually applying it comes with feature loss and other problems. ZFS vs encryption is a bad place to live. You of course need encryption for your VMs on the Enterprise.
Proxmox is a lovely piece of software, highly recommend to everyone. But I think Proxmox and vSphere target rather different scales of cluster. Proxmox starts to choke once your past 20 nodes or so. Corosync just wasn't designed for huge clusters.
I think the price hikes have made the "medium" use-case more uncertain.
I know VMWare works great for a lot of people, but it was always just a thorn in my side. Almost a decade ago my small company moved from VMWare to ganeti, and we've been really happy with it.
At the time we were running a SAN and the small VMWare Essentials license. But managing it with Windows was painful for me personally, and our HP SAN was kind of painful, every year like clockwork it would just stop processing until we had remote hands reboot it. And every year they'd tell us "Try this firmware" to shut us up for another year, until eventually the product reached EOL.
We had been running ganeti in our dev/stg cluster for years at that point (and the python.org infrastructure was using it at that time as well), so I decided to transition us to local storage running VMWare and then to ganeti. Doing local RAID array monitoring on VMWare was quite painful at the time, maybe that's gotten better?
I absolutely love that ganeti has such a good suite of command-line tools for operating. I have a small shell script that respins half our dev/stg infrastructure every other day. Our entire critical infrastructure is HA at the machine level, so we don't play around with anything like live VM moves, we just do offline moves. I used to do the drbd-based replication and live moves, but the doubling of disk writes and committed disc usage wasn't worth it for us.
Proxmox is on my radar if we needed to move away from ganeti, but so far we are looking like we'll stick with ganeti.
The latest release is almost 3 years ago, but the repo has commits a month ago. So I guess it depends on your definition of "alive". As "stable" software, I'm fine with it not having a ton of active development. It's been a real workhorse for me, I've got it running on 2 clusters of 6 physical machines each (1.4TB total RAM, 50TB disc, each).
~6 years ago I went to a local Dell dog and pony show where someone at the next table was poo-pooing on "those greybeards that run Debian and ganeti" instead of VMWare. But, this shit works...
So Broadcom is basically just milking VMware until it completely dies? I guess you should always consider the risk of a company buying a product you use and then doing this
VMWare has been dying a death of 1000 cuts for years. It's been passed around to now the third owner. They created an entire market and have been watching it erode away.
Yeah but how does it make any business sense for Broadcom to be doing this? They could have raised prices 100% and fewer people would flee and they'd make more money overall. 1,000% price raises are just plain extortion.
If the action of a company doesn't seem to make sense, ask yourself: could you imagine an incentive structure within the company that could lead to such actions?
Mathematically if they can extract say 1.5 times the money they used to buy something in 3 years. It is a good deal for them. And even better if they can extract more money beyond those 3 years and then sell what is still left to reclaiming bit more money. Making 15% a year on investment and it fully paying back is not bad move.
Did Broadcom think the VMware moat was a lot deeper than it is? Or do they think losing small or medium customers is worth it to get 10x more revenue from its largest customers?
I'm curious why anyone still uses the product. They had a clear advantage 20 years ago but that was 20 years ago. I really don't understand what distinguishing thing they offer these days.
Not having a clear advantage might be a good reason to not choose it for new products, but there are thousands of companies that chose VMware 20 years ago when it did have a clear advantage. You have to have a pretty good reason to invest thousands of people hours in a project to migrate to another plaform, and simply "it's not the latest product out there" is not often a compelling reason to throw out something that still works pretty well.
Price increases as in the linked article can be a compelling reason, of course.
If a company has a significant amount of infra set up with VMware the costs of migration to another solution are substantial. They have to retrain staff, who may not be keen on relearning how to do what they already know how to do with VMware. They have to replace other solutions that work with VMware but maybe not a new alternative. And there’s usually features that VMware has that others do not, or they work differently enough that it introduces more friction.
The core virtualization stuff is commodity, but there’s much more to it.
A lot of people knows it very well. I wouldn't consider it for personal use but I didn't have a lot of issue managing vsphere clusters when I was still into it.
Also it is not about the VMs only but the whole ecosystem. Last time I worked for a company where we considered moving out of vmware, the main issue was not migrating to a new VM management solution but having to migrate to a new VM hypervisors + migrating to a new backup solution supporting the new hypervisor. This while still maintaining original backup solution until retention runs out or we migrate and test all old backup with long retention. Or we had to decide on maintaining several backup systems in parallel which would increase the cost of backups.
VMware was worth the money because it just worked and worked decently. I chose it in the past to avoid wasting my time on crashes, incompatibilities, and lacking performances that were typical for other VMs.
A lot of people use it. I for one use it because it is simple to manage, setting up shared storage is a breeze, live migration is super easy. I don't use any of the fancy NSX stuff and whatnot because I don't believe they add any value.
I was looking at Proxmox, but I kind of got lost in the documentation when it comes to shared storage and high-availability, meaning restarting VMs on another host if one host fails.
RedHat is just as expensive as VMware used to be, but it had less features, so it is not really an option.
Broadcom is the typical MBA run company. Nobody fucking cares about the long run. The idea is to extract as much as much profit as possible right now, the executives pocket the bonuses, and when the well dries, they can just move to the next target.
Broadcom kind of went against this though as it also divested itself of several parts of the vmware product range. A lot of what we used to see advertised alongside sphere isn't even a broadcom product now.
> do they think losing small or medium customers is worth it to get 10x more revenue from its largest customers?
they'd be almost certainly right to think that, of course it'd be worth it. The pending question is whether the largest customers will actually sticks with VMWare for them to actually see this 10x revenue materialise
I would guess they know it's dying and they're trying to get everything they can out of it while they can. Vendor lock-in is a real risk everyone should be cognizant of.
Readit News
In the end I left before broadcom bought vmware but the status at the time was "let's wait a bit more".
However Red Hat now only supports conversions to OSV, since RHV was deprecated (sadly).
The biggest problem is oVirt itself has not proven to be a very sustainable open source project. If oVirt dies, we'll likely remove support in v2v. (There's a great start up opportunity here, for a dull but money-making company that productizes oVirt again.)
To move VMs from RHV to OSV you can just copy the disk image since they already should have virtio drivers, qemu guest agent, and be able to boot on any qemu/KVM-based platform. I believe there's some automation for that, but it doesn't involve virt-v2v.
6 months ago on HN: https://news.ycombinator.com/item?id=40452260
> Steve McDowell, chief analyst at NAND research, told The Register that VMware by Broadcom is “laser focused on high-revenue, high-margin business” and has priced its wares “just below the pain threshold for customers they care about.”
They did exactly that, and the sales manager was fired over losing one of the biggest clients and the loss of revenue that came with it.
One needs to be very, very careful about both the pain threshold and the "** you, I'm moving my business elsewhere" threshold.
But if Broadcom increases price 10-fold and loses half the customers by revenue over the next couple of years, that reflects in their bonus and they can then move on.
The customers might have a plan to move off vmware by 2027, but continue to pay the money until then. Come the 2027 renewal the execs that made the massive bonuses for increasing revenue have left and then when the company says they aren't renewing it's the next generation dealing with the problems, and then they just reach for the three-envelopes.
This really is on the companies that choose a single supplier. If you had two providers, then shifting doesn't take a 3 year program, it's just a policy change to prefer Provider B rather than Provider A.
In sweden it's really hard to find a pub that has coca cola, they almost exclusively have pepsi.
Also in the supermarket pepsi is cheaper.
A segment of a customer base is hardly a buffer, if you underserve them they will leave, and you will have wasted effort making an inclusive sales and support process to them.
Deleted Comment
Dead Comment
https://www.ciodive.com/news/broadcom-att-vmware-settlement-...
From what I have seen they are just a PE firm pretending to be a tech company. They will bleed the customers who can’t move fast enough. Burn it to the ground to get every last dollar.
Broadcom aren't hiking the price 10x for everyone.
If you previously used:
Then you're going to see you vSphere costs jump massively because you're only option to buy vSphere now also comes with: So, if you're already heavily invested in VMWare, you might even see your costs reduce.A sale isn't free. It requires time, headcount, and engineering+PM resources.
At Broadcom size, you want to maximize sales to F1000 customers (the kind that can and will shell 7-8 figures in TCV) as that gives you the most bang for your buck from a SalesOps perspective.
A major issue with VMWare was their sales team would chase logos at the expense of an optimal sale, so their account teams would give extreme discounts (70-80%) at barely above the margin in order to close deals.
Chances are if you're a vmware estate, those products are going to integrate better and then you only have a supplier, support portal, customer rep, etc..
Broadcom loose customers who arn't really invested in to VMWare and those who stay become even more invested in to them.
Deleted Comment
What I picked up is that before the acquisition, price was mostly determined by CPUs on the hypervisor, now it's mostly by RAM.
Which means we now have to optimize hypervisor hardware to a different licensing model.
I don't think we've seen 10x price hikes, but it could possible if a provider has optimized hard for a metric that is now less relevant, and/or has difference licensing deals.
Caveat: I've picked all that up second-hand through the grapevine, I'm not working with our virtualization farms directly. So not 100% on top of what's going on.
I think you might have that backwards, before we paid based on ram and now we pay based on cores
Didn't even oracle drop this monstrosity of a licensing fee? Didn't microsoft also dropped / stoped pursuing it? How on earth do people justify this?
I think the price hikes have made the "medium" use-case more uncertain.
At the time we were running a SAN and the small VMWare Essentials license. But managing it with Windows was painful for me personally, and our HP SAN was kind of painful, every year like clockwork it would just stop processing until we had remote hands reboot it. And every year they'd tell us "Try this firmware" to shut us up for another year, until eventually the product reached EOL.
We had been running ganeti in our dev/stg cluster for years at that point (and the python.org infrastructure was using it at that time as well), so I decided to transition us to local storage running VMWare and then to ganeti. Doing local RAID array monitoring on VMWare was quite painful at the time, maybe that's gotten better?
I absolutely love that ganeti has such a good suite of command-line tools for operating. I have a small shell script that respins half our dev/stg infrastructure every other day. Our entire critical infrastructure is HA at the machine level, so we don't play around with anything like live VM moves, we just do offline moves. I used to do the drbd-based replication and live moves, but the doubling of disk writes and committed disc usage wasn't worth it for us.
Proxmox is on my radar if we needed to move away from ganeti, but so far we are looking like we'll stick with ganeti.
~6 years ago I went to a local Dell dog and pony show where someone at the next table was poo-pooing on "those greybeards that run Debian and ganeti" instead of VMWare. But, this shit works...
[1] https://www.netcials.com/financial-revenue-history-usa/11246...
In order to sell?, spinoff?, or use the cash to finance something else?
Price increases as in the linked article can be a compelling reason, of course.
The core virtualization stuff is commodity, but there’s much more to it.
Also it is not about the VMs only but the whole ecosystem. Last time I worked for a company where we considered moving out of vmware, the main issue was not migrating to a new VM management solution but having to migrate to a new VM hypervisors + migrating to a new backup solution supporting the new hypervisor. This while still maintaining original backup solution until retention runs out or we migrate and test all old backup with long retention. Or we had to decide on maintaining several backup systems in parallel which would increase the cost of backups.
It is not just about migrating VMs.
Inertia, both in a company and in the industry, is a major reason.
I was looking at Proxmox, but I kind of got lost in the documentation when it comes to shared storage and high-availability, meaning restarting VMs on another host if one host fails.
RedHat is just as expensive as VMware used to be, but it had less features, so it is not really an option.
It is the locust economy.
They are just not selling products individually anymore.
If you previously only licensed a single product, your only option now is to buy everything under a single license.
We are already using a lot of VMWare products, not only do we now have access to more, our overall price went down.
I think medium sized customers have a choice, switch from their Broadcom products, or move more to Broadcom.
they'd be almost certainly right to think that, of course it'd be worth it. The pending question is whether the largest customers will actually sticks with VMWare for them to actually see this 10x revenue materialise