AT&T Claims VMware By Broadcom Offered It a 1,050% Price Increase (theregister.com) 48
The Register's Simon Sharwood reports: AT&T has claimed that Broadcom made it an offer to increase prices by 1,050 percent, and may be influencing other vendors to make a migration harder. The claim of the colossal price hike came in an email [PDF] filed in evidence by AT&T in its case alleging Broadcom hasn't honored a contract that would allow the carrier to acquire an additional two years of support services for its VMware estate. The email was penned by AT&T executive vice president and general manager Susan A Johnson and appears to be addressed to Broadcom CEO Hock Tan.
"After a 10 plus year strategic relationship with Broadcom ... I am sad to report that we appear to be at an impasse on our VMware deal," Johnson wrote on August 19. "The latest offer that we have received would put us at an average of $REDACTED per year for a 5 year deal, where we currently pay $REDACTED per year to support previously purchased perpetual licenses with a right to renew support through September, 2026. This proposed annual increase of +1,050 percent in one year is extreme and certainly not how we expect strategic partners to engage in doing business with AT&T."
"After a 10 plus year strategic relationship with Broadcom ... I am sad to report that we appear to be at an impasse on our VMware deal," Johnson wrote on August 19. "The latest offer that we have received would put us at an average of $REDACTED per year for a 5 year deal, where we currently pay $REDACTED per year to support previously purchased perpetual licenses with a right to renew support through September, 2026. This proposed annual increase of +1,050 percent in one year is extreme and certainly not how we expect strategic partners to engage in doing business with AT&T."
Migrate (Score:4, Insightful)
Re: (Score:2)
Broadcom's entire business strategy is to bet this isn't the case (and make enough money off their locked-in customers, who have to pay 10x, to make up for the inevitable loss of those customers eventually).
close (Score:5, Insightful)
Jackpot! (Score:4, Insightful)
That's their plan. Jackpot it then sell the husk at a discount but keep the patents.
Re: (Score:1)
Re:Migrate (Score:5, Insightful)
The fancier versions of VMware have features that those platforms don't.
At least not yet. A lot of companies are going to do the math and realize it makes more sense to pay a few developers -- either in-house or freelance -- to develop the FOSS alternatives rather than pay Broadcom's rates.
Re: Migrate (Score:5, Interesting)
Re: (Score:1)
10 years ago, that was definitely all possible to do in KVM, hell, networked USB and PCIe fabrics were introduced in Linux as soon as USB/PCIe became a thing. VMware is running on a Linux kernel, so the hardware abstraction is done in Linux.
Re: Migrate (Score:5, Interesting)
Re: Migrate (Score:1)
clustering, health checking, and automatic failover
Seriously, Proxmox was introduced in 2008, RHV in 2009, OpenStack 2010, they are all fully automated VM clusters with things like live migration, replication etc
Re: (Score:2)
I wish Red Hat kept with RHEV, rather than just dump it for OpenShift. Not everyone needs Kubernetes, and Kubernetes has a lot of issues if using it for persistent, "pet" VMs. RHEV/RHV was very competitive.
OpenStack is pretty much dead, unless it is used internally by a company that has thrown a LOT of manpower to keep it up. For example, if Keystone hangs or glitches in any way, kiss your overcloud/undercloud goodbye.
I'm guessing the best thing is probably Proxmox or XCP-ng, although Proxmox is pretty n
Re: (Score:1)
Xen is supported by Citrix. OpenShift still runs VMs, Nutanix is a big 'proprietary' KVM vendor (they just don't tell you). Dell has some branded solution, Ubuntu has 'white glove' KVM clusters.
Re: (Score:2)
>"Getting there today with the likes of Proxmox."
And XCP-NG/Xen Orchestra
https://xcp-ng.org/ [xcp-ng.org]
https://en.wikipedia.org/wiki/... [wikipedia.org]
https://xcp-ng.org/blog/2023/1... [xcp-ng.org]
Re: (Score:2)
Proxmox was doing all 3 10 years ago. It leveraged pacemaker originally.
Re: (Score:1)
I guess you guys have never heard of Proxmox?
Re: (Score:1)
Re: Migrate (Score:1)
Proxmox has resource pools which you can group not just VMs, and you can switch the view to use them in the GUI, I see references going back to 2009.
will ATT can pay for them to add stuff that is mis (Score:3)
will ATT can pay for them to add stuff that is missing but they also need to keep VMWare alive till they have something new to move to.
Re: (Score:2, Informative)
Such as ...
Tell me a feature that you can't find in both Xen (Citrix, XCP-ng...) and KVM (Red Hat, Ubuntu, Proxmox, Nutanix, ...) - because quite a lot of them are now offering direct conversion as a service. Just point it at your VMWare API or storage pool and start migrating.
Conversely, VMware for the past decade or so has been missing or delaying features that are available in KVM. Things like SecureBoot, vTPM, VFIO, DPU, high-end Ethernet, object storage, Kubernetes, VMware eventually came with a half-a
Re: (Score:2)
VMWare has five things that put it ahead of everyone else:
It has an awesome, federatable, control plane. It is relatively easy to manage a large cluster and data center of ESXi nodes. Nothing out there really has this much scalability.
VMFS is the best there is. It is just plug and play. No setting witness volumes, no secondary connections... just if the ESXi nodes can communicate with each other and the shared LUNs, it "just works". This is definitely not the case with Hyper-V.
VMWare Fault Tolerant VMs
Re: (Score:2)
ZFS is definitely the king, but btrfs is used in a lot of areas. All new Synology models use it by default. For RAID, they put btrfs on md-raid, which ensures a solid block level layer that doesn't have RAID-5/RAID-6 write hole issues, and put btrfs on top of that, or on top of LUKS.
Another nice thing it has is deduplication which doesn't add RAM. ZFS and kmod-kvdo require approximately a gig of RAM for every terabyte deduplicated. btrfs's deduplication is offline, which means one runs a tool like duper
Re: (Score:2)
You also forgot that it had excellenent support and a vast network of resellers, independent consultants, VARs, and so on, except that Broadcom nuked all that Day 1.
Re: (Score:2)
The other important thing to consider is that, despite the price going up by over 1000%, the number of developers at "VMWare" actually working on "VMWare" is rapidly approaching zero. They've been slashing that team f
Re: (Score:2)
Every employee is a liability. Every customer is a liability. Get rid of as many of those as you can. Good numbers must go up, bad numbers must go down.
Re: (Score:3)
This wasn't a serious comment was it? VirtualBox is a desktop hypervisor with no clustering, HA, or fault tolerance features.
QEMU is a system emulator and can be a core component of (and often is) a virtualization offering like Proxmox, RHEV, OpenStack, OpenShift, etc. but by itself is not suitable for replacing an entire stack of technologies, which is what VMWare vSphere is.
Broadcom is a shitty company but vSphere probably is the the most mature and feature-filled player in virtualization right now.
Re: (Score:3)
QEMU is a system emulator and can be a core component of (and often is) a virtualization offering like Proxmox, RHEV, OpenStack, OpenShift, etc. but by itself is not suitable for replacing an entire stack of technologies, which is what VMWare vSphere is.
Broadcom is a shitty company but vSphere probably is the the most mature and feature-filled player in virtualization right now.
The issue is KVM requires work / knowledge to do this. vmware dumbed it down to a windows admin level. And of course the execs all went oooh shiny. Combining network / storage / vm into a single place simplified the ownership of vmware, which broadcom is abusing the customer base for.
Re: (Score:2)
No, because Virtual Box and Qemu don't have features like clustering, high availability and vMotion.
Enterprise organisations aren't just installing VMware on a single box and spinning up a handful of virtual machines on it.
Re: (Score:2)
Re: (Score:2)
arguably for this particular market segment, they'd be more likely to be talking to RedHat , Proxmox or even Microsoft. Virtual Box is cool and all, but I'm not sure if its got the hooks and performance to be that useful to large datacenters
Re: (Score:2)
I mean, I think virtual box and q-emu are pretty mature at this point, has to be cheaper to migrate and leave broadcom with decreasing market share
Aside from these not being comparable to VMware's offerings in the enterprise space, this takes time. Which is what Broadcom is doing: Extract as much money as absolutely possible from VMware in the short term, knowing that in a couple of years the offering is dead and the customers are gone. You don't increase prices by 1050% - or even an order of magnitude less - if you want to keep your customers and relationships for the long term.
Looks like the shoe is on the other foot. (Score:5, Funny)
Terrible. AT&T is becoming the victim of a large, monopolistic technology firm. Won't someone please come to their aid?!
Re: (Score:3)
Terrible. AT&T is becoming the victim of a large, monopolistic technology firm. Won't someone please come to their aid?!
Yeah, it's always nice when the hunter becomes the hunted or, as in this case, when the fucker becomes the fucked. Gotta love it when the rich eat their own.
Re: (Score:2)
While I understand the sentiment, who do you think is going to pay the price in the end?
Re: (Score:3)
While I understand the sentiment, who do you think is going to pay the price in the end?
Customers too lazy to switch carriers.
Ya, but ... (Score:5, Funny)
AT&T Claims VMware By Broadcom Offered It a 1,050% Price Increase
Maybe that's for Broadcom's Ultra Unlimited Premium VMware Plan* ! :-)
[* Speeds may be throttled to Unlimited Starter Plan speeds during periods of high demand, 250 VMs, and/or after 1TB of monthly data usage.]
(What's good for the Goose ... )
how AT&T deals with partners (Score:4, Funny)
This proposed annual increase of +1,050 percent in one year is extreme and certainly not how we expect strategic partners to engage in doing business with AT&T."
Huh... Isn't that how they tend to treat -customers-?
Rate hikes suck (Score:3)
Sounds like my DirecTV NOW increase. Heh.
Windows 365 (Score:2)
AT&T is a slow learner (Score:5, Insightful)
This is something AT&T fails to figure out over and Over and OVER again.
( Software licenses, hardware vendors and even contract labor. They simply can't figure this out. )
When you switch to and / or rely upon a third party vendor to do anything for you, the initial price tag is going to be a fantastic deal.
You can't possibly pass it up. The cost savings are just amazing. Bonuses for everybody ! LOL
As time goes on, and you begin to rely more and more on the aforementioned third party, the price starts to creep up a bit but, not too much.
Once a large portion of your business relies on this vendor, they set the hook and can now dictate the prices you WILL pay if you want your
business to keep running smoothly. Of course you can always switch to another vendor, but it will be a major pita and will take not only years
to do but an enormous amount of funding / manpower to do it.
The folks who got you hooked on their product know this and exploit the living hell out of it.
Re: (Score:2)
That said, I feel little sympathy for ATT. My i
Re: AT&T is a slow learner (Score:2)
AT&T is arguing that they have a contract that entitles them to extend legacy support at specific prices at their option for a specific period, and Broadcom unilaterally declared that it was null and void, just because they said so.
I'm not a fan of AT&T, but they seem to me to have the better argument.
Not hard enough (Score:2)
Looks like Hock isn't hocking ATT hock enough.
Oracle. (Score:3)
Taking lessons out of Oracle's book, I see.
Trying to extract value before K8s kills them (Score:2)
VMware probably sees the writing on the wall w/ Kubernetes. It's an absolute game changer compared to running heavy VMs for every little business app.
Sure, it's overkill for small businesses, but any business big enough to need a large fleet of "enterprise VMs" can just rack up a bunch of bare metal, slap Kubernetes and Rancher on there and reap the benefits.
Open Source (Score:1)