But all of these moves have been with the goals of innovating faster
Fail.
meeting our customers’ needs more effectively,
Fail.
and making it easier to do business with us.
Fail.
If the goal is milking your customers for more money then it all makes sense. At least until they start migrating away from VMware and your new client stream completely dries up.
Executives care only about their bonuses, and the best way to boost them is to milk companies on the short term even if it ruins them on the long term. There is no incentive to care about long term.
At least until they start migrating away from VMware and your new client stream completely dries up.
This is a feature. They take a product that you can’t easily move off of, like VMware or Symantec. They jack up the prices. A few people leave, but some people struggle to leave and pay lots of money. Demand drops, they cut most of the staff and put the product on life support, and eventually they kill it. By that time they’ve bought another company to do the same.
It’s what we always did with cows. To have a dairy cow you need to repeatedly get it pregnant. Why keep feeding an older less productive cow that has already calved 4 or 5 times when you’ve already got yourself a few new heifers.
fyi most dairy cows are kept beyond the age of 3. You can see on the World Dairy Expo YouTube channel that there are cows producing up to the age of 6.
Heifers reach puberty at about 12 months old, but some delay until up to 20 months old. After breeding they gestate for 283 days (almost 9 1/2 months), give birth to their calf and give milk for about 10 months. “Culling” a cow at the age of three would mean only getting 2 full lactation cycles of milk from them, which isn’t super economical.
You’re not too far off the mark, though: once any individual animal becomes less profitable than a new one could be, they get culled.
Said it before, they’re dumping low value customers for high value customers. This can be a legit strategy!
But not in this case. As you said, they’re going to milk the cow dead. We wouldn’t touch VMware with a ten-foot frog. Leaning into Proxmox ATM, working great so far.
While open shift can somewhat accommodate VM workloads, it feels like an afterthought. Really the goal is to charge how your applications are run.
However, VMware presides over staunchly “old fashioned” ships that think in terms of “machines”. So ProxMox is a bit more similar.
RedHat did have ovirt which would have been a closer platform, but they ditched that in favor of openstack, which was also VM centered but “cloudy”, which also isn’t the target model of on premise virtualization (openstack had other problems too), and now it’s openshift, which is largely a “kubernetes is a buzz word, let’s go, also as an afterthought some VM hosting to give some semblance of continuity for users we yanked through RHEV, openstack, and for now openshift”
It might play a role, but ProxMox may be better situated to be like for like. Microsoft is of course pushing their azure stack for those wiling to get tied up into azure a bit. I suspect openshift will continue to mainly focus on cloud hosted VMs rather than retool they’re go to market to better capture those abandoning VMware. After all, since the story is “reduce costs”, that’s not an appealing scenario to red hat/IBM, since it inherently puts an obvious upper bound on revenue and the customers will be those that demonstrated they are the most ready to migrate when unhappy.
I think this is very interesting. I’ve been investigating the solutions and found that a lot of other platforms like scale hci, virtuozzo, and oracle VM are all based on the oVirt/Openstack platforms but have been customized.
Openshift and Suse Harvester both have a very similar Kubernetes first approach which I think is interesting. Harvester seems to rely on KubeVirt to deploy “legacy workloads” (probably windows).
The reason I mention openshift though is because I’ve been paying attention to Wendell, level1techs, and the level1 forums and Wendell keeps hinting that Redhat/IBM openshift + intel is being used as a VDI platform featuring Intel flexGPUs for a secret customer (I wouldn’t be surprised if it was a government facility like the national laboratory near Knoxville). I’m just trying to envision how that deployment looks.
Fail.
Fail.
Fail.
If the goal is milking your customers for more money then it all makes sense. At least until they start migrating away from VMware and your new client stream completely dries up.
Yes, the only thing that Broadcom gives a shit about here:
The ‘ol late-stage capitalism adage “growth above all else”
“Short-term growth above all else, then lay everyone off and find the next business to buy out and ruin”
Executives care only about their bonuses, and the best way to boost them is to milk companies on the short term even if it ruins them on the long term. There is no incentive to care about long term.
Investors reward this behavior. The system is designed to destroy good things.
This is a feature. They take a product that you can’t easily move off of, like VMware or Symantec. They jack up the prices. A few people leave, but some people struggle to leave and pay lots of money. Demand drops, they cut most of the staff and put the product on life support, and eventually they kill it. By that time they’ve bought another company to do the same.
They’re basically just milking old cows to death.
Totally unrelated, but that’s actually what we’re doing with cows. Killing them of at age 3 bc they don’t give enough milk anymore
It’s what we always did with cows. To have a dairy cow you need to repeatedly get it pregnant. Why keep feeding an older less productive cow that has already calved 4 or 5 times when you’ve already got yourself a few new heifers.
fyi most dairy cows are kept beyond the age of 3. You can see on the World Dairy Expo YouTube channel that there are cows producing up to the age of 6.
Heifers reach puberty at about 12 months old, but some delay until up to 20 months old. After breeding they gestate for 283 days (almost 9 1/2 months), give birth to their calf and give milk for about 10 months. “Culling” a cow at the age of three would mean only getting 2 full lactation cycles of milk from them, which isn’t super economical.
You’re not too far off the mark, though: once any individual animal becomes less profitable than a new one could be, they get culled.
Yeah, im not that big on the details, since I don’t support those industries at all.
Said it before, they’re dumping low value customers for high value customers. This can be a legit strategy!
But not in this case. As you said, they’re going to milk the cow dead. We wouldn’t touch VMware with a ten-foot frog. Leaning into Proxmox ATM, working great so far.
A lot of people talk about Proxmox but I wouldn’t be surprised if redhat open shift decided to throw its hat into the ring as an HP or Dell partner.
While open shift can somewhat accommodate VM workloads, it feels like an afterthought. Really the goal is to charge how your applications are run.
However, VMware presides over staunchly “old fashioned” ships that think in terms of “machines”. So ProxMox is a bit more similar.
RedHat did have ovirt which would have been a closer platform, but they ditched that in favor of openstack, which was also VM centered but “cloudy”, which also isn’t the target model of on premise virtualization (openstack had other problems too), and now it’s openshift, which is largely a “kubernetes is a buzz word, let’s go, also as an afterthought some VM hosting to give some semblance of continuity for users we yanked through RHEV, openstack, and for now openshift”
It might play a role, but ProxMox may be better situated to be like for like. Microsoft is of course pushing their azure stack for those wiling to get tied up into azure a bit. I suspect openshift will continue to mainly focus on cloud hosted VMs rather than retool they’re go to market to better capture those abandoning VMware. After all, since the story is “reduce costs”, that’s not an appealing scenario to red hat/IBM, since it inherently puts an obvious upper bound on revenue and the customers will be those that demonstrated they are the most ready to migrate when unhappy.
I think this is very interesting. I’ve been investigating the solutions and found that a lot of other platforms like scale hci, virtuozzo, and oracle VM are all based on the oVirt/Openstack platforms but have been customized.
Openshift and Suse Harvester both have a very similar Kubernetes first approach which I think is interesting. Harvester seems to rely on KubeVirt to deploy “legacy workloads” (probably windows).
The reason I mention openshift though is because I’ve been paying attention to Wendell, level1techs, and the level1 forums and Wendell keeps hinting that Redhat/IBM openshift + intel is being used as a VDI platform featuring Intel flexGPUs for a secret customer (I wouldn’t be surprised if it was a government facility like the national laboratory near Knoxville). I’m just trying to envision how that deployment looks.
I find it peculiar that everybody in this discussion is ignoring Hashicorp’s stuff??
https://en.wikipedia.org/wiki/HashiCorp
I glanced at Proxmox vs Hashicorp’s stuff, after seeing some discussions on here about 'em, & Hashicorp’s stuff is oriented to clarity & simplicity.
Sorta Japanese take on it.
To me clarity is worth a significant amount of value.
Anyways, I’m just noting this, for anyone who’s actually considering such things: I’m only a nobody who has avoided geeking for some years.
( :
I never thought of hashicorps products as a replacement for VMware but more of an add on to it.