- cross-posted to:
- news
- cross-posted to:
- news
Since Broadcom’s $61 billion acquisition of VMware closed in November 2023, Broadcom has been charging ahead with major changes to the company’s personnel and products. In December, Broadcom began laying off thousands of employees and stopped selling perpetually licensed versions of VMware products, pushing its customers toward more stable and lucrative software subscriptions instead. In January, it ended its partner programs, potentially disrupting sales and service for many users of its products.
This week, Broadcom is making a change that is smaller in scale but possibly more relevant for home users of its products: The free version of VMware’s vSphere Hypervisor, also known as ESXi, is being discontinued.
You must log in or register to comment.
Proxmox my beloved
Did you ever try LXD/Incus?
Nope. I’m pretty happy with proxmox and I dont want to change a perfectly fine, running system
+1 … been using PVE in my homelab for ages and just deployed a small, self-contained (i.e. non-SAN-connected) PVE cluster at the office in light of Broadcom’s shenanigans. I had no idea just how fantastically well Proxmox ran on higher-end hardware with Ceph installed. It’s glorious.
Oh yeah it runs fine until they kill their free tier like ESXi did or… it completely fails over and over again.
Proxmox doesn’t have a free tier, it is free.
You can pay for support and shit if you want.
Since you are apparently on an anti-proxmox crusade. Have you tried that iscus thing in enterprise? Like a very large scale production deployment? Since I have never heard of it, I am curious if anyone dares to use it in enterprise when people are even scared of proxmox or anything not VMware or MAYBE hyperV
Since you are apparently on an anti-proxmox crusade. Have you tried that iscus thing in enterprise? Like a very large scale production deployment?
Maybe if you read the comment I linked you’ll find that that’s precisely what we had with Proxmox and then migrated to LXD.
I am curious if anyone dares to use it in enterprise when people are even scared of proxmox or anything not VMware or MAYBE hyperV
I guess it depends on the kind of “enterprise” we’re talking about. If your “enterprise” is somewhat of a provider / ISP they should be okay with LXD. A lot of service providers are running some form of LXC/LXD right now with very good results.
If by “enterprise” you mean your typical 400+ people company that does something not related to tech with an overworked and barely competent IT / infrastructure team, then the answer is: they won’t move out of vmware ever.
Well, suit yourself. :)
I love LXD/INCUS + ZFS! It goes together like chocolate and peanut butter.
Oh yeah, zfs send ftw. I personally run most of stuff on BTRFS and I can say the same.
Once you’ve throughly beaten your head against every little thing that’s not ready to go out of the box like ESX is, puzzled through cryptic VM errors and Ubuntu being broken on default VM settings, and then browsed the sometimes aggressively unhelpful forums, it’s great!
sounds like a skill issue to me
sounds like a skill issue to me
Ah yes, I see the forums are leaking again
You are either going all in with VMware or you’re dead to them. Full suite or nothing, take your pick.
The moment that broadcom bought them the writing was on the wall. Many people have already jumped ship.
A lot of people can’t jump ship, at least not within a year or two.
To what?
I’ve got a client who is currently a vmWare shop that (along with moving datacenters) is migrating to hyper-v when they rebuild.
I hope you mean Azure Stack HCI, seeing how Hyper-V 2019 is the end of the line.
I do. we’ve already deployed it internally once, and will be deploying additional clusters over the coming year.
How do you like it so far? I’ve got a few customers interested.
It’s alright, but it really isn’t my favorite. We spun up the cluster using professional support services from our vendor and it was rocky af, and the built in dashboard reporting is worthless if you want to know what’s been provisioned instead of straight utilization. Alerting has been another struggle for us as well.
I’m sure it would work better when more integrated with azure, but for our 100% local workload it leaves a lot to be desired. But thankfully since it’s windows based and manageable with powershell I was able to write a custom report to surface the metrics my teams and management care about.
Perhaps I’ll try it out for a while before making such a huge commitme… oh, i see…
For the cost, SMB is going to walk away. There are millions of SMB’s.
SMBs are not the target. Companies with a sizeable vSAN investment, huge amounts of VMware based automation and the fortune 1000 are. MSRP on the cheap license is going to be around $275/core, minimum 16 cores per socket.
And a potential 90% discount for big customers?
Probably not that deep. I’ve heard there are definitely discounts. That doesn’t count for much though when it still increases your cost 6x.
Shockingly, no.
That’s simply short sighted.
So they ignore the Fortune 1000+1 (the up-and-coming 1000). They also stop providing a learning/familiarity path.
I’m already seeing SMBs looking at KVM, Proxmox, Xen, etc. When these young engineers/managers/architects grow and move to Enterprise, what are they going to recommend when VMware is $300/core?
I’m all for (as in I push) recognizing the value of (even expensive) licensing when it reduces engineering costs and complexity, but that’s what I’d call a “metric shitload”.
A mid-size business could easily justify transitioning to just about any other VM solution when faced with that kind of increase. One 16-core host is now $5k in licensing, practically doubling the cost - and that’s an annual cost for years - saddling “future IT” with that cost that can now no longer be invested elsewhere.
Now imagine you have 10 such boxes.
They are planning to tolerate losing 95% of their customers. Of about 100,000 customers, they only care about 600 of them much, and about 6 thousand kind of, if they want to stick around, but not too much. The rest are fully expected to bail.
https://www.theregister.com/2022/05/30/broadcom_strategy_vmware_customer_impact/
LOL was about to implement esxi, on a rather beefy surplus server, to run all my students’ PCs on since win 11 won’t boot on their hardware from 24h2… Guess my students won’t get to use VMware and the purchase approval I just got for a few workstation pro licenses wasn’t needed.
Proxmox for baremetal hypervisor, or? I’ve got a bunch of windows server licenses as well, I think some for hyper-v server as well. What would you implement?
I’m happy with proxmox in a non-production environment/homeLab. Stable and straightforward.
Just found out from your comment that windows is shutting the door completely on CPUs that don’t support POPCNT. There’s config settings to install Windows 11 on legacy hardware (old CPU, tpm chips, etc) but who knows when they’ll pull the plug on that.
Weird as it is, it’s not as radical as I thought:
For Intel’s chips, it was added as part of SSE4.2 in the original first-generation Core architecture, codenamed Nehalem. In AMD’s processors, it’s included in SSE4a, first used in Phenom, Athlon, and Sempron CPUs based on the K10 architecture. These architectures date back to 2008 and 2007, respectively.
Of course they probably could have avoided it, but a 15 year old PC is as close to ewaste as it gets. Even if you could run Linux on it, a modern smartwatch probably has more computing power, let alone a smartphone or raspberry pi. The main use could be as a space heater.
Yea there’s plenty of reasons to shit on MS, but dropping support for 15+ years old CPUs isn’t one of them lol
If someone is being affected by this, then maybe it’s time for them to upgrade their shit lmao
Proxmox is really good, same with XCP-ng. You could also run something like Debian server and roll your own KVM based platform if you have the chops.
Overall, lots of solid choices in the Open Source realm. I would avoid proprietary solutions, since that’s largely the reason the whole VMWare situation happened in the first place.
Check xcp-ng with xen-orchestra.
Here’s a really nice guide to XCP-NG vs Proxmox (Video creator’s preference is for XCP, so there’s an acknowledged bias there, but it’s still a solid rundown of the two).
Personally, I just run straight KVM on Debian or Ubuntu servers, but that’s not for everyone. Web based management for KVM is still kind of rough. Cockpit is getting there, but it’s missing key features, and the web based graphical console absolutely sucks.
Here is an alternative Piped link(s):
Piped is a privacy-respecting open-source alternative frontend to YouTube.
I’m open-source; check me out at GitHub.
I really like XCP-ng. Imo the interface is more understandable and polished than Proxmox. Similar to vSphere + vCenter, but more advanced options as well.
Hyper-V or nutanix community. those will be the dominant hypervisors in the near future. I can see nutanix really taking off soon if they cant reach some of the features that VMware had. Hyper-v is sort of stuck since their host OS layer sucks, but it’s also pretty cheap.
They’ve “redone” hyper-v a bit with the Azure HCI stack, but yeah, the OS layer sucks. Even more so than normal windows server.
I know internally Broadcom is screwing over VMware employees with new contracts. I’ve heard of staff pushing back three times so far to get contracts changed.
They also poached all of the biggest customers from a bunch of their resellers, from what I hear. I’m guessing the reseller market is pretty dead.
This is exactly what happened.
Well shit. ESXi was the best way to build a home lab when studying for the professional certifications I need.
Proxmox?
Proxmox is a good option for home labs (in my opinion) but it sucks if your workplace utilizes VMWare (or a product limited to VMWare) and you want practice at home
I haven’t worked with it before. The product is only supported on VMWare hypervisors, so no matter what, I’ll have to build on an unsupported setup, but I was leaning towards KVM for familiarity. I will make sure to check Proxmox out too though.
Check out XCP-ng. Open source, enterprise grade bare metal hypervisor.
I moved from ESXi to it about a year ago, it’s been solid. Lots of documentation and support from the community. Lawrence Systems has a ton of great videos on configuring it, both simple and advanced.
Which enterprise is using XCP-ng?
Large portions of IBM, Rackspace, Alibaba, Oracle, and AWS’s cloud infrastructure are powered by the Xen Project hypervisor, which is the core of the XCP-ng stack.
Why can’t we have nice things
We do, we have Proxmox.
And LXD/Incus…
Why use Proxmox when half of it’s technology (the container part) was made by the same people who made LXD/Incus? I mean Incus is free, well funded and can be installed on a clean Debian system with way less overhead and also delivers both containers and VMs.
To be honest, I’ve never heard of Incus before… it does look interesting though. Terraform provider too? I will check it out!
In a word? Capitalism
You can have nice things… for money!
Yes… for price set by the greed of companies
Or for effort…
And everyone clapp… oops not that kind of story
And everyone saw that coming
And everyone saw that coming
I wonder where all the “wE jUSt GoTTa WAiT aNd SEe” people are now?
This is the best summary I could come up with:
In December, Broadcom began laying off thousands of employees and stopped selling perpetually licensed versions of VMware products, pushing its customers toward more stable and lucrative software subscriptions instead.
This week, Broadcom is making a change that is smaller in scale but possibly more relevant for home users of its products: The free version of VMware’s vSphere Hypervisor, also known as ESXi, is being discontinued.
ESXi is what is known as a “bare-metal hypervisor,” lightweight software that runs directly on hardware without requiring a separate operating system layer in between.
ESXi allows you to split a PC’s physical resources (CPUs and CPU cores, RAM, storage, networking components, and so on) among multiple virtual machines.
ESXi also supports passthrough for PCI, SATA, and USB accessories, allowing guest operating systems direct access to components like graphics cards and hard drives.
It was also a useful tool for people who used the enterprise versions of the vSphere Hypervisor but wanted to test the software or learn its ins and outs without dealing with paid licensing.
The original article contains 334 words, the summary contains 172 words. Saved 49%. I’m a bot and I’m open source!
