Hacker News Reader

Comment by sofixa | Hacker News Reader

It's embarrassing how poorly Broadcom have rushed into throwing the baby out with the bathwater for no reason. They could have taken a slower approach instead of doing everything as soon as the deal officially closed...

I hope they lose as much business as possible over this, everyone should be looking into alternatives because they're such a shitshow.

View on HN

Yeah, I don't get why they just did the Elon thing of slash and burn without any plan. I get that they want to upend the business model, focus on whatever end of the market, etc. Why not do that in a planned out, measured way? Why confirm literally every concern people had when the acquisition was announced? Why get yourself into what will obviously be a ton of lawsuits over people who can't use the product they already paid for?

Why wait? Just do it and deal with the ill effects as soon as possible.

This way they get to slash costs with less staff and effectively fire the clients they don’t want.

They expect backlash and thinks it’s worth the cost, because they plan on milking the remaining customers asap.

But they're cutting off customers who have paid and are under contracts

I'm not talking about ripping off the bandaid on the new sales channels, I'm talking about the fact that customers who have active licenses they've paid for can no longer activate them because they've shut these systems down

That inevitably is going to lead to a class action against them

They have budgeted for lawsuits and don’t care. It’s a feature not a bug to get rid of clients who won’t be milked to the bone.

I am suspicious that they even thought about it that far. There's no way it's cheaper to let the active customers sue them rather than just leave the license activation servers up

How does that make sense? Had they chosen a slow gratuitous path those customers might've switched to other products, now they've been burned and will think twice before choosing VMware again.

Vmware is dead. They don't plan to acquire new customers (or invest a cent in research& development). The purpose of this whole thing is to milk those fortune 500 companies that are too entrenched to migrate away.

Ahh the grand BigCo tradition of acquiring a company and then setting fire to it.

> They expect backlash and thinks it’s worth the cost, because they plan on milking the remaining customers asap.

Why would they think that after that disaster said customers will still be with them? I no longer have the displeasure of being a VMware customer, but if I saw what they're currently doing while jacking up prices and trimming support, what possible reason would there be to remain as a customer? Inertia? At some point you know the bill will be too high, so might as well move before it gets too bad.

VMware will end up like Symantec or similar, a shell of a company that people know of, but nobody actually uses on purpose.

That's the whole plan. They basically only want to retain the fortune 500 customers and wing them dry. Anybody that can migrate away will do so, but the rest will be taken to the cleaners.

There's little competition in some of VMWare's segments and even if there is, the migration is long, expensive and risky.

Most of the damage was already done during acquisition process (simply based on Broadcom's reputation). There simply isn't any incentive to deviate from the original plan to milk the customers.

>everyone should be looking into alternatives

I'm interested to hear this community's take on what an alternative to vmware is. And I'm not talking about just ESX, I'm talking about their entire ecosystem including vsphere and all of it's features that an enterprise environment depends on.

No two environments are the same. Instead of looking for a like for like total replacement (even though you actually don't use all VMware features or use some only because they were the only way possible in VMware's ecosystem and not because they're actually any good (vRA, Log Insight)), evaluate your actual needs and what do you need to run under what conditions. You'll probably end up discovering that any of Proxmox, OpenStack Nutanux, oVirt, Ganetti, Kubernetes (with or without KubeVirt), Nomad (with or without VMs) will cover most if not all that you need.

You will also probably end up discovering your costs drastically inflate. There is a reason why companies stick with bundled products. Find a way to bundle that mismash of software you mentioned into a single pane of glass with a 4 year support guarantee and you will be rich.

You will quickly find out why it hasn’t been done and why Redhat and Microsoft look so promising after all

Ha, nonsense. VMware products are notoriously super expensive. From discussions with people at organisation switching to alternatives, their cost are going down 30-50%, with all the added benefits of a proper modern orchestrator instead of an obsolete hardware simulator. (Stuff like having a way of securely introducing secrets or security authenticating workloads, or integrated deployments, healthchecks, etc.)

Also, "bundled" stuff constraints you. You're forced into using VMware's crap orchestrator or log management tool because it's the one that works best with the mess that is other VMware products.

OpenStack

So 10 years ago I was part of a project which pitted VMware vsphere against openstack. VMware was cheaper opex by a factor of 4 and they were using ESXi as the hypervisor. Openstack is a non starter for enterprise environments.

Now with containers and kubernetes being the standard, openstack is another very very heavy distribution your organization must maintain

> VMware was cheaper opex by a factor of 4 and they were using ESXi as the hypervisor

How could it be 4x cheaper in Opex when OpenStack is an open source project with multiple competing vendors available if you want support? In terms of non-licensing costs, OpenStack has a few more services that need to be deployed separately, but I don't see that costing 4x, especially when those services often don't have VMware equivalents.

OpenStack's got Magnum for that need.

The framework for deployment has vastly improved.

Reply on news.ycombinator.com