• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Supermarket Giant Tesco Sues VMware, Warns Lack of Support Could Disrupt Food Supply

Apparently we managed to make Broadcom mad at us and they no longer wants our business since we did not want to do a "give os ALL your hypervisor business" so mangement has no other option than to follow alongside my plan about stopping doing business with them meaning my timetable just got accelerated.
No wonder people are leaving their products on droves and it is going to be very interesting to see their financials in 2028/2029, as the increase in license fees no longer can cover up the shrinking customer base.
 
A lot of the smaller guys are now only given the option for a 1 year license opposed to a 3 year. So I have the pleasure of re-IPL to HyperV for all of the ESXi systems. Much cheaper since they already have the CAL's for windows. Broadcom really eff'd a lot of our customers. But they don't seem to care because about 85% of their customer base is the big guys that have to pay because redoing everything to another hypervisor would be way too much money and effort.
 
Edgar said:
So I have the pleasure of re-IPL to HyperV for all of the ESXi systems.
Click to expand...

I actually like and prefer Hyper-V
1772669487918.png
 
kram182 said:
I actually like and prefer Hyper-V
Click to expand...
I will say I find managing issues easier since it has a full OS under it. Back when we used ESX (which granted was well over a decade ago) it was a real shitshow the one time there was a big issue since the local console had almost zero capabilities.

Also it seemed to me that Hyper-V was much easier to set up in a non-clustered but redundant setup. Useful when you don't have a lot of infrastructure, as in our case, where a cluster would just be more cost.
 
kram182 said:
Hyper-V itself is still bare-metal/type 1 though (OS is secondary/side-car)
Click to expand...
I should clarify that I mean for management. I know that when you run Hyper-V is actually takes over and the host OS runs virtualized, but it still functions as the host and you can use it to manage the system. With ESX all management was doing online which was fine if online was working, but if your DNS servers were in your hypervisor, as ours were, and it all went to shit... well their console was extremely lacking.

The fact that I can just RDP or do a PSSession to a Hyper-V server and admin it like that is nice for me. It acts like Hyper-V is running on top of that, even if that isn't true in the strict sense.

It also has plenty of remote management tools, of course.

But all our Hyper-V servers are wired to a Raritan KVM so if shit hits the fan, I can get in and do things with it directly.
 
Edgar said:
A lot of the smaller guys are now only given the option for a 1 year license opposed to a 3 year. So I have the pleasure of re-IPL to HyperV for all of the ESXi systems. Much cheaper since they already have the CAL's for windows. Broadcom really eff'd a lot of our customers. But they don't seem to care because about 85% of their customer base is the big guys that have to pay because redoing everything to another hypervisor would be way too much money and effort.
Click to expand...
Just and wait and see when the "bigger guys" do a hardware refresh as were my original plan but since Broadcom decided to absolut muppets we will start to migrate before our hardware gets refreshed.
Chasing away customers seldom ends well and when datacenter/infrastucture guys gets burned we never forget I can sya I know Vmware if off the table as an option going forward for me, even if I should get a better offer and change job/company and I saw an article where 73% of the polled VMware customers were actively exploring alternatives.
 
Our HyperV poc is mostly finished and we are moving forward with the migration over the next 4 months. The one thing that concerns me is managing Windows updates on the hosts themselves. VMware updates were fairly infrequent, but hyperV is going to require monthly windows patching, and we all know how that has been going lately. To add the the complexity, my director cut our host count down to the point the total memory usage of all VMs will barely fit under the remaining hosts if we take one down for patching or other concerns.

How do some of you already on HyperV handle the monthly updates?
 
littlegreenmenfromhome said:
Just and wait and see when the "bigger guys" do a hardware refresh as were my original plan but since Broadcom decided to absolut muppets we will start to migrate before our hardware gets refreshed.
Chasing away customers seldom ends well and when datacenter/infrastucture guys gets burned we never forget I can sya I know Vmware if off the table as an option going forward for me, even if I should get a better offer and change job/company and I saw an article where 73% of the polled VMware customers were actively exploring alternatives.
Click to expand...
Agreed. It will bite them in the butt later down the line for sure. I hope it does. They screwed so many customers. My company is actively looking for other hypervisor solutions to be compatible with on our enterprise hardware (currently support ESXi, HyperV, RHEL and Oracle). I suggested proxmox since it is gaining a lot of steam lately on the consumer side.
 
Edgar said:
Agreed. It will bite them in the butt later down the line for sure. I hope it does. They screwed so many customers. My company is actively looking for other hypervisor solutions to be compatible with on our enterprise hardware (currently support ESXi, HyperV, RHEL and Oracle). I suggested proxmox since it is gaining a lot of steam lately on the consumer side.
Click to expand...
The cluster size (and needing a 3rd party for support outisde business hours) is why we cannot go Promox on everthing, our large HCI clusters will most likely go the Nutanix route.
Also using this an an oppertunity to consolidate our infrastructure (location wise) and cut license cost (with Proxmox on non-customer enviroments and smaller internal clusters).
 
  • Like
Reactions: Edgar
like this
littlegreenmenfromhome said:
The cluster size (and needing a 3rd party for support outisde business hours) is why we cannot go Promox on everthing, our large HCI clusters will most likely go the Nutanix route.
Also using this an an oppertunity to consolidate our infrastructure (location wise) and cut license cost (with Proxmox on non-customer enviroments and smaller internal clusters).
Click to expand...
I think my companys solution is better than nutanix. But I am biased :) . Let me know if you want more info. You mightve heard of us already.
 
Edgar said:
I think my companys solution is better than nutanix. But I am biased :) . Let me know if you want more info. You mightve heard of us already.
Click to expand...
I am always open to suggestions, one issue though is that we run a lot of Cisco UC applications and (just like with HyperFlex) a one stop vendor like Nutanix makes a lot of sense.
 
Cactusj said:
How do some of you already on HyperV handle the monthly updates?
Click to expand...
Depends on what it running on it. For a couple of our servers, we just do it. They are down for like 5 minutes max and usually less, that kind of downtime is perfectly tolerable for the stuff that runs on it at night. There really aren't that many workloads that REALLY demand 24/7/365 uptime and the VMs themselves get patched once a month (be they Linux or Windows). For stuff that really does need better uptime, the best answer is redundancy in VMs that are spread across the hosts. Again, the hosts can go down for patching whenever, and the redundant VMs handle everything.

However for stuff that can't be (or more accurately the researchers won't spend the time to) made redundant then the answer is failover or a cluster. Failover is pretty easy to set up, you just have two hosts and set up replication between the two of them. That's a good idea to do for backup anyhow. Then when you want to patch the primary, you tell the VMS to fail to the backup, then fail them back. It is all live migration. If you are willing to spend more money, you do a clustered setup. You have multiple hosts in a cluster, connected to shared storage (which of course needs to be reliable, I'ma recommend NetApp as I always do). At that point you can move the VMs around freely and basically instantly and if a host dies they'll just pop up on another host. Requires more hardware, of course.
 
Cactusj said:
Our HyperV poc is mostly finished and we are moving forward with the migration over the next 4 months. The one thing that concerns me is managing Windows updates on the hosts themselves. VMware updates were fairly infrequent, but hyperV is going to require monthly windows patching, and we all know how that has been going lately. To add the the complexity, my director cut our host count down to the point the total memory usage of all VMs will barely fit under the remaining hosts if we take one down for patching or other concerns.

How do some of you already on HyperV handle the monthly updates?
Click to expand...
Sycraft said:
Depends on what it running on it. For a couple of our servers, we just do it. They are down for like 5 minutes max and usually less, that kind of downtime is perfectly tolerable for the stuff that runs on it at night. There really aren't that many workloads that REALLY demand 24/7/365 uptime and the VMs themselves get patched once a month (be they Linux or Windows). For stuff that really does need better uptime, the best answer is redundancy in VMs that are spread across the hosts. Again, the hosts can go down for patching whenever, and the redundant VMs handle everything.

However for stuff that can't be (or more accurately the researchers won't spend the time to) made redundant then the answer is failover or a cluster. Failover is pretty easy to set up, you just have two hosts and set up replication between the two of them. That's a good idea to do for backup anyhow. Then when you want to patch the primary, you tell the VMS to fail to the backup, then fail them back. It is all live migration. If you are willing to spend more money, you do a clustered setup. You have multiple hosts in a cluster, connected to shared storage (which of course needs to be reliable, I'ma recommend NetApp as I always do). At that point you can move the VMs around freely and basically instantly and if a host dies they'll just pop up on another host. Requires more hardware, of course.
Click to expand...
The monthly updates are a deal breaker for me/my work. Some at my organization were pushing straight up HyperV and I had to explain that, with various pfsense (and other critical networking) running as VMs in our cluster, you're going to have some outages if the host reboots every few weeks (without major changes management wasn't interested in implementing anyways...).

That then turned them onto the path of Azure HCI (now "local"), another pain in the ass for unrelated reasons. Following that long move, the results were a mixed bag - with HCL lists proving to be inaccurate, driver issues, and a very shoddy local administration tools. Long story short, management was again displeased (this time with HCI), and they finally threw up their hands and paid for us to build out another VMWare cluster. So we've come full circle back to VMWare and now I expect Proxmox is likely on the next project.
 
SamuelL421 said:
The monthly updates are a deal breaker for me/my work. Some at my organization were pushing straight up HyperV and I had to explain that, with various pfsense (and other critical networking) running as VMs in our cluster, you're going to have some outages if the host reboots every few weeks (without major changes management wasn't interested in implementing anyways...).
Click to expand...
I mean... I would argue that everyone should be updating almost everything at least monthly these days. If the downtime is intolerable, then engineer it redundant such that there isn't downtime. Not just for security reasons but because if you can't tolerate brief downtime for updates, how are you going to tolerate downtime for something going wrong?

We do have things, like our storage, that we don't tolerate downtime on and hence it is redundant.
 
Sycraft said:
I mean... I would argue that everyone should be updating almost everything at least monthly these days. If the downtime is intolerable, then engineer it redundant such that there isn't downtime. Not just for security reasons but because if you can't tolerate brief downtime for updates, how are you going to tolerate downtime for something going wrong?

We do have things, like our storage, that we don't tolerate downtime on and hence it is redundant.
Click to expand...
If you do not plan redundancy or plan for "Murphy" you are designing to fail indeed.
 
littlegreenmenfromhome said:
Apparently we managed to make Broadcom mad at us and they no longer wants our business since we did not want to do a "give os ALL your hypervisor business" so mangement has no other option than to follow alongside my plan about stopping doing business with them meaning my timetable just got accelerated.
No wonder people are leaving their products on droves and it is going to be very interesting to see their financials in 2028/2029, as the increase in license fees no longer can cover up the shrinking customer base.
Click to expand...
Reality is, the big customers they care about are likely getting great service, or enough, and are SO en-grained in their tech stack that to move out would cost them millions and millions and years to migrate and Broadcom knows this. So they will keep milking it for a couple more years likely. And if any of the larger clients do leave, so be it, they can prob take enough of a hit and will then just extract more profits from existing until they make back their money, then the product will slowly tank..
 
littlegreenmenfromhome said:
If you do not plan redundancy or plan for "Murphy" you are designing to fail indeed.
Click to expand...
Exactly, but trying to sell double the cost to some companies is literally impossible as they do not see the value...

Until the day things go sideways...then it is "why is this down! How did this happen!" and you resend the emails about why you needed x2 of everything..
 
MrGuvernment said:
Exactly, but trying to sell double the cost to some companies is literally impossible as they do not see the value...

Until the day things go sideways...then it is "why is this down! How did this happen!" and you resend the emails about why you needed x2 of everything..
Click to expand...
Yup been there done that and i always make sure that the client understand that they got exactly what they paid for and then hope they learned their lesson
 
littlegreenmenfromhome said:
Yup been there done that and i always make sure that the client understand that they got exactly what they paid for and then hope they learned their lesson
Click to expand...
It's this and the entrenched technical debt in the form of "well it's already setup and it works, why would we pay more?" - Why? because your shit has zero redundancy and you keep paying us to slap band-aids on the problem.
 
SamuelL421 said:
It's this and the entrenched technical debt in the form of "well it's already setup and it works, why would we pay more?" - Why? because your shit has zero redundancy and you keep paying us to slap band-aids on the problem.
Click to expand...
I have found that sometimes a simple "risk letter" suddenly motivates people as it seems they do not want to be the one left holdning the "monkey" why Murphy says "hi".
 
  • Like
Reactions: Edgar
like this
You must log in or register to reply here.
Back
Top