Hock Tan doesn’t really know anything about software
He’s a financial vampire. But deliberately limiting your number of customers while keeping the products on life support is insane.
Broadcom has signaled its $61 billion acquisition of VMware will involve a “rapid transition from perpetual licenses to subscriptions.” That's according to Tom Krause, president of the Broadcom Software Group, on Thursday's Broadcom earnings call. He was asked how the semiconductor giant plans to deliver on its guidance that …
We were thinking of transiting to VMware, but that is obviously totally out of the question now. We're a small lab, our budget doesn't allow us to pay (through the nose) over and over again for the same infrastructure.
Oh well, it's not like it will hurt them. They will be able to milk those who can't quit to cover all losses, and reduce their support expenses by pruning the smaller, less profitable clients like us.
Tell your customers that 'we the corporation' are going to gouge more £££/$$$$ etc.
See Rule 1
Don't the schmucks know that business confidence is dropping?
Trying to get more blood out of a stone is not going to go down well with customers.
I'll be making this change known to my bosses this morning. They won't be happy as we use VMware a lot in testing. It looks like it is back to Virtualbox even though I suspect that Oracle will see this and do exactly the same...
As regards the LibreOffice database I agree with you.
However I get the impression that desktop database products as a whole are aimed at those who have finally realised the shortcomings of spreadsheets as databases and contrive to reassure by looking as much like spreadsheets as possible. Kex looks more like Access than does LO Base but from an old Informix hand this is damning with faint praise. In particular I'd prefer the approach of "dump controls for the columns of a single from the table(s) I've selected on a new form, provide a menu control for CRUD*, forward & backward and let me take things from there". And I wish they wouldn't devise their own database engines; just use Sqlite for a strictly one user, local data source or ODBC to link to any well-known engine.
* Create, Retrieve, Update, Delete.
you only buy VmWare because you want to
You also buy it when you need a relatively easy to install and run bare-metal hypervisor, and can't afford to hire some IT whizz guy who could assemble and maintain a working system out of open source solutions.
We know open source solutions exist, because we actually use them right now. But when you don't have money for a dedicated IT department, beyond a given level you're forced to try to simplify things by going for commercial solutions: The half a day you wasted browsing through contradicting online forums for a solution is half a day you didn't do the actual work you're paid for. Sad, but that's life.
Our product is a turnkey system based around a single decently powered server. What we pay for a perpetual ESXi license on that server is about what an IT guy would bill us in a month. But here's the fun part: Our product goes into secure locations on private, air-gapped networks. We absolutely cannot be put in a position where VMware has to phone home periodically to check if the bill's been paid.
I remember when they started.
Hardware failed, new hardware requires new OS that requires new app software that requires new drivers that don’t support anything u have.
Lots of churn.
Hardware fails, it is a VM so move to another piece of hardware. Done. No churn.
That was worth something, but the cat is out of the bag and we have choices.
The no churn is Bad for the industry. So force it:
1. Buggy software, churn
2. Security issues (see #1), churn
3. “ end of life /support”, churn
4. And the latest - “supply chain issues” , churn
Careful with "Alternatives" - think scale - we're still running some 700 VMs across two sites with site to site failover - ProxMox is KVM/Containers underneath - it's the management toolset that makes it useful, but to what scale?
So main players are VMware/Microsoft Hyper-V/SCVVM (ugh) /OpenStack (Complex) I guess - though VMware probably has most clever tricks in the bag. VMware's main competitor now has to be Azure/AWS?
oVirt. The upstream (opensource) free version of Red Hat Virtualization (RHV). KVM under the hood.
It mostly follows the VMware vCenter/ESXi model, e.g. oVirt has the Engine (front end for command and control, like vCenter), and Nodes (hypervisor systems, like ESXi).
Based on Linux, typically CentOS or RHEL, which figures since it's Red Hat-sponsored. Personally I'd prefer Alma or Rocky these days. Or, really, Debian, but if you want that Linux family under the covers your best bet might be Proxmox.
I played with both oVirt and Proxmox a while ago in the lab, both were decent. Nice enough front-ends with GUI and VM console etc. like you'd expect. Again, KVM doing the hypervisor work.
For small labs (e.g. 1 hypervisor, maybe a dozen VM's or so) I would probably use Proxmox these days, or stock KVM+qemu+libvirt and stitch together the associated extras I might want.
For multi-hypervisor server setups I'd check out oVirt. I don't know what Red Hat's RHV $$ version gets you beyond oVirt, presumably paid support; I've never looked into it, merely assumed it's like RHEL from a licensing and support etc. standpoint.
Not to the user. Lifetime cost of availability will go though the roof. Plus there's the permanent sword of Damocles in case a subscription renewal fails. I've personally witnessed this with certificates, and Equifax lost the decryption certificate on their IDS for the same reason, leading to one of the largest data breaches on record.
The bean counters take command yet again. IT is no longer there to support the user - it's there to maintain the vendor's revenue stream. I wonder why the phrase 'protection racket' keeps flashing through my mind.
I've been trying to buy a simple license upgrade for a Symantec security appliance, and not only has Broadcom been mind-blowingly incompetent, they've been *actively hostile* and stubbornly refuse to lift a finger to solve the problem they've created. (Okay, they've lifted the middle finger.) This was also a transition from perpetual to subscription licensing, and it's turned my H/W product into a brick in lieu of a $$$$ payment covering the whole company.
Brocade has still been okay, but that's an old-line H/W business.
This post has been deleted by its author
"Krause also repeatedly said Broadcom .... is pleased to be acquiring a sales organization and channel relationships that give it reach Broadcom does not currently enjoy."
They flushed the sales organization & channel relationships which came along with Symantec. Maybe they are now realizing how important that is to the software business?
Capex vs opex - companies are often happier with ongoing opex. They are probably paying a yearly support fee anyway, so it's just how the pricing breaks down.
There are a number of public clouds that depend on VMWare - if the subscription pricing is sensible, this actually encourage using it within a data centre.
There are already other virtualisation products that have a subscription pricing model - some of them based on Open Source technologies.
Rancher has nothing to do with KVM and SUSE hasn't done a lot for KVM either (you can't even get a supported version of Cockpit for SEL or openSUSE Leap, although for some reason they do one for rolling release openSUSE Tumbleweed).
RHEL and derivates (Rocky/Alma/Oracle Linux) make a much better virtualization host out of the box than SUSE. So does Ubuntu, actually.
"Whether it's perpetual or subscription, frankly, it's the same,"
Is it indeed? Well if it is then it'll be unlike every other subscription service, which end up costing more than a perpetual license when spread over the lifespan of the equipment it's installed on. Not to mention, of course, the continual stream of new 'features' which we don't want and which sometimes break other things (Microsoft's specialism).
There is big challenges & opportunities for Broadcom. VMware is overpriced today and many organizations don't see a good ROI. So Broadcom will have to adjust VMware's pricing model if they hope to gain & retain market share. Now that said I don't think a subscription model is necessarily a bad thing for Vmware. Doing so could potentially callow customers more flexibility to choose the products & features they need. The other challenge is to show companies the value proposition for VMware. The potential is there. But the cost is simply to high for many organizations. Private cloud management is probably the real opportunity for VMware.
Biting the hand that feeds IT © 1998–2022