If all hypervisors were priced the same...
-
@tim_g said in If all hypervisors were priced the same...:
@olivier said in If all hypervisors were priced the same...:
@tim_g said in If all hypervisors were priced the same...:
If features and costs (free) were identical across the board, I would choose KVM hands down.
I love being able to run off Fedora Server, plus all the doors that open up by doing that... which you can't get from Hyper-V or VMWare.
Sure Xen can be installed on there too, but it's dieing and I'm less familiar with it.
Can you stop with that FUD? Thanks. It's not dying at all. I hear this since 2006. It's like saying Linux is not secure because Open Source.
No fear or doubt, just uncertainty. But this is only because of how Citrix is treating Xen Server, and how Amazon is moving from Xen to KVM.
I feel the only thing that can save Xen is XCP-ng. I'm really hoping for it's success and have high hopes for it.
That's because you have a very partial view of Xen project. Xen project is far more than XenServer/XCP. Xen is the core hypervisor, used by a LOT of companies (from automotive to the Cloud).
A lot of companies are using it Xen + their own toolstack without making publicity around it (like AWS, which is NOT leaving Xen, just adding some instance on another HV to get some specific features not in Xen yet). Some companies (Gandi) even switch from KVM to Xen:
https://news.gandi.net/en/2017/07/a-more-xen-future/
So your opinion is mainly forged by limited number of sources, in a loop of telling "Xen is dying" since 10 years. The main reason is that because Xen is far less "segmented" than KVM (eg: easier to make clickbait articles on Xen security issues than KVM, despite KVM sec process is almost catastrophic/non-transparent)
-
@olivier said in If all hypervisors were priced the same...:
@tim_g said in If all hypervisors were priced the same...:
@olivier said in If all hypervisors were priced the same...:
@tim_g said in If all hypervisors were priced the same...:
If features and costs (free) were identical across the board, I would choose KVM hands down.
I love being able to run off Fedora Server, plus all the doors that open up by doing that... which you can't get from Hyper-V or VMWare.
Sure Xen can be installed on there too, but it's dieing and I'm less familiar with it.
Can you stop with that FUD? Thanks. It's not dying at all. I hear this since 2006. It's like saying Linux is not secure because Open Source.
No fear or doubt, just uncertainty. But this is only because of how Citrix is treating Xen Server, and how Amazon is moving from Xen to KVM.
I feel the only thing that can save Xen is XCP-ng. I'm really hoping for it's success and have high hopes for it.
That's because you have a very partial view of Xen project. Xen project is far more than XenServer/XCP. Xen is the core hypervisor, used by a LOT of companies (from automotive to the Cloud).
A lot of companies are using it Xen + their own toolstack without making publicity around it (like AWS, which is NOT leaving Xen, just adding some instance on another HV to get some specific features not in Xen yet). Some companies (Gandi) even switch from KVM to Xen:
https://news.gandi.net/en/2017/07/a-more-xen-future/
So your opinion is mainly forged by limited number of sources, in a loop of telling "Xen is dying" since 10 years. The main reason is that because Xen is far less "segmented" than KVM (eg: easier to make clickbait articles on Xen security issues than KVM, despite KVM sec process is almost catastrophic/non-transparent)
I see. That makes sense.
-
@tim_g said in If all hypervisors were priced the same...:
@olivier said in If all hypervisors were priced the same...:
@tim_g said in If all hypervisors were priced the same...:
If features and costs (free) were identical across the board, I would choose KVM hands down.
I love being able to run off Fedora Server, plus all the doors that open up by doing that... which you can't get from Hyper-V or VMWare.
Sure Xen can be installed on there too, but it's dieing and I'm less familiar with it.
Can you stop with that FUD? Thanks. It's not dying at all. I hear this since 2006. It's like saying Linux is not secure because Open Source.
No fear or doubt, just uncertainty. But this is only because of how Citrix is treating Xen Server, and how Amazon is moving from Xen to KVM.
Amazon is a concern, but Citrix is not, IMHO. Citrix has been out to cripple Xen for years and, if anything, it just shows a lack of health of Citrix, not Xen.
-
Once Xen gets the PV driver features backported to core Xen PV, we will see a leap forward too, I think.
-
@scottalanmiller I don't see exactly what are you talking about. What's PV driver feature?
-
@scottalanmiller Citrix doesn't care anymore on server virt market, since a while now.
-
@olivier said in If all hypervisors were priced the same...:
@scottalanmiller Citrix doesn't care anymore on server virt market, since a while now.
Did they ever? They bought Xen for the name so that they could confuse their customers into thinking that XenApp was somehow virtualization.
-
@olivier said in If all hypervisors were priced the same...:
@scottalanmiller I don't see exactly what are you talking about. What's PV driver feature?
Xen has some performance advantages using their PV drivers over doing full PV.
-
@scottalanmiller You should have mixed some stuff. PV mode doesn't need PV drivers by definition. You meant HVM? (to be in PVHVM then?)
-
@olivier said in If all hypervisors were priced the same...:
@scottalanmiller You should have mixed some stuff. PV mode doesn't need PV drivers by definition. You meant HVM? (to be in PVHVM then?)
I know it doesn't but it needs the performance tech from them.
-
@olivier said in If all hypervisors were priced the same...:
@tim_g said in If all hypervisors were priced the same...:
If features and costs (free) were identical across the board, I would choose KVM hands down.
I love being able to run off Fedora Server, plus all the doors that open up by doing that... which you can't get from Hyper-V or VMWare.
Sure Xen can be installed on there too, but it's dieing and I'm less familiar with it.
Can you stop with that FUD? Thanks. It's not dying at all. I hear this since 2006. It's like saying Linux is not secure because Open Source.
Do you have any 3rd party surveys or tracking showing growth in Xen, because all the public (and private sets like IDC) that I’ve seen show It loosing market share.
-
@scottalanmiller said in If all hypervisors were priced the same...:
@olivier said in If all hypervisors were priced the same...:
@scottalanmiller Citrix doesn't care anymore on server virt market, since a while now.
Did they ever? They bought Xen for the name so that they could confuse their customers into thinking that XenApp was somehow virtualization.
They bought it because Vmware bundled the hypervisor with their VDI product, so Citrix bought Xen and had its devs focus on VDI friendly features (APIs for provisioning, and GPU support). They briefly tried to take on ESXi in the enterprise but abandoned that a few years back.
Citrix also pushed cloudstack for a while to hosting providers (but seems to have given up on that too).
-
@storageninja said in If all hypervisors were priced the same...:
@scottalanmiller said in If all hypervisors were priced the same...:
@olivier said in If all hypervisors were priced the same...:
@scottalanmiller Citrix doesn't care anymore on server virt market, since a while now.
Did they ever? They bought Xen for the name so that they could confuse their customers into thinking that XenApp was somehow virtualization.
They bought it because Vmware bundled the hypervisor with their VDI product, so Citrix bought Xen and had its devs focus on VDI friendly features (APIs for provisioning, and GPU support). They briefly tried to take on ESXi in the enterprise but abandoned that a few years back.
Citrix also pushed cloudstack for a while to hosting providers (but seems to have given up on that too).
I doubt that that is why they bought it. That makes no sense since they could have done THAT without buying it. They bought it for the name alone.
-
I would split those depending on the infrastructure size like:
- Small business with only 2-3 hosts would be completely fine with Hyper-V
- For a larger one with up to 10 hosts I would prefer going with VMware
- For everything larger KVM or XEN. Probably KVM because of larger community and better self-supporting options...
-
@net-runner said in If all hypervisors were priced the same...:
I would split those depending on the infrastructure size like:
- Small business with only 2-3 hosts would be completely fine with Hyper-V
- For a larger one with up to 10 hosts I would prefer going with VMware
- For everything larger KVM or XEN. Probably KVM because of larger community and better self-supporting options...
My question here is... what makes Hyper-V or VMware better on those small systems? Hyper-V's main problems, mostly huge management overhead and complexity, are worst at the small scale. where KVM or Xen's ease of use is a really big deal.
VMware I see in that mid-range... but companies in that range are crippled by cost today. If it was free, I think it would make sense all the way down. Hyper-V really depends on "free" more than KVM and Xen do.
-
@scottalanmiller said in If all hypervisors were priced the same...:
Once Xen gets the PV driver features backported to core Xen PV, we will see a leap forward too, I think.
Didn't Amazon shift everything away from PV because of security? (There are a LOT fewer instance types of PV these days).
Other hypervisors moved away from PV for computing a long time ago as VT-x and newer hardware functionality (PCID etc) simply made the juice not worth the squeeze.
-
@scottalanmiller said in If all hypervisors were priced the same...:
My question here is... what makes Hyper-V or VMware better on those small systems? Hyper-V's main problems, mostly huge management overhead and complexity, are worst at the small scale. where KVM or Xen's ease of use is a really big deal.
VMware I see in that mid-range... but companies in that range are crippled by cost today. If it was free, I think it would make sense all the way down. Hyper-V really depends on "free" more than KVM and Xen do.There are quite a few points but one low hanging fruit is the DRS family of features (Compute/network DRS, Affinity rules, Storage DRS, SIOCv2 VAIO filters, Proactive - DRS). It's balancing logic is significantly more advanced. combined with better scheduler overheads, more advanced new workload placement logic means you can get by with a lot less hardware.
For someone with 10 tiny VM's this isn't going to matter, but for someone who's operating with a decent amount of scale having to throw money at hardware, and bodies instead of software become a trade-off that throws things into DRS being worth the premium for TCO.
Now if the hardware is free to you, and labor is $2 an hour then TCO will shift the other way vs paying for software.
Also, decisions are often more nuanced than simple TCO decisions. If you have compliance requirements this often shifts to commercial solutions that have validated FIPS 140-2 modules/solutions. If you need a DISA STIG at a given level paying some money and being able to deploy a single VIB to harden compliance vs. go through checklists and argue with auditors can be a big deal. How do you quantify the cost of applying with NIST for validation with a do it yourself setup vs. a turnkey solution?
The cost of management tools are generally looked at as a function of the cost of existing management labor (People), the cost of the solution stack, and the premium for availability.
If you have Oracle RAC or SQL Always ON clusters that cost 40K per host in licensing it's different math. Paying 2K for some hypervisor management tools that will let you run 1.5x to 2.5x denser on host usage (and drop associated licensing costs), or free up 15% time for a Sysadmin who's paid 100K so he can go get other projects finished, isn't a "Crippling cost" but a simple, logical conclusion.
Customers who need VM Fault Tolerance don't care what the cost is because the alternative is generally proprietary solutions that cost 250K per server, or death (US wrongful death is what 2.5 million each?) or re-writing their application and getting it revalidated by regulators(Millions in capex if even an option).
If you have Excel/Access Databases and 5 Windows XP VMs, and you have outsourced your sysadmin work to SouthEast Asia for 5K a year, and an outage is going to cost you nothing sure.
-
@storageninja said in If all hypervisors were priced the same...:
@scottalanmiller said in If all hypervisors were priced the same...:
My question here is... what makes Hyper-V or VMware better on those small systems? Hyper-V's main problems, mostly huge management overhead and complexity, are worst at the small scale. where KVM or Xen's ease of use is a really big deal.
VMware I see in that mid-range... but companies in that range are crippled by cost today. If it was free, I think it would make sense all the way down. Hyper-V really depends on "free" more than KVM and Xen do.There are quite a few points but one low hanging fruit is the DRS family of features (Compute/network DRS, Affinity rules, Storage DRS, SIOCv2 VAIO filters, Proactive - DRS). It's balancing logic is significantly more advanced. combined with better scheduler overheads, more advanced new workload placement logic means you can get by with a lot less hardware.
For someone with 10 tiny VM's this isn't going to matter, but for someone who's operating with a decent amount of scale having to throw money at hardware, and bodies instead of software become a trade-off that throws things into DRS being worth the premium for TCO.
That's what I meant. I totally get that stuff at scale, even a little scale. But the point there was on "tiny" systems that he was thinking VMware at one size and Hyper-V at another, but because things were really small.
-
@storageninja said in If all hypervisors were priced the same...:
Also, decisions are often more nuanced than simple TCO decisions. If you have compliance requirements this often shifts to commercial solutions that have validated FIPS 140-2 modules/solutions. If you need a DISA STIG at a given level paying some money and being able to deploy a single VIB to harden compliance vs. go through checklists and argue with auditors can be a big deal. How do you quantify the cost of applying with NIST for validation with a do it yourself setup vs. a turnkey solution?
RHEL/RHV have a good solution here. Auditors go through OpenSCAP scans with nice HTML reports and we justify any “failures.” It’s a pretty nice system. You can even scan live running VMs without the agent with KVM. It uses libguestfs tools and mounts the guest read only to scan the guest.
-
@stacksofplates said in If all hypervisors were priced the same...:
Also, decisions are often more nuanced than simple TCO decisions. If you have compliance requirements this often shifts to commercial solutions that have validated FIPS 140-2 modules/solutions. If you need a DISA STIG at a given level paying some money and being able to deploy a single VIB to harden compliance vs. go through checklists and argue with auditors can be a big deal. How do you quantify the cost of applying with NIST for validation with a do it yourself setup vs. a turnkey solution?
RHEL/RHV have a good solution here. Auditors go through OpenSCAP scans with nice HTML reports and we justify any “failures.” It’s a pretty nice system.
That just audits if it was set. What I'm talking about is a single package you deploy that goes ahead and sets the configuration settings up for you.
On ESXi you can use Update Manager to track compliance with the DISA VIB, and use that for tracking it. Just attach as a baseline to your clusters and let Update Manager keep it up to date. Ed Groggin I think has a tool that will do an auto-generation of a report on the hardening guidelines.
Looking online, I'm not seeing Server 2016 in STIG viewer yet. Has Microsoft not gotten a STIG out yet?
Also Redhat Virtulization licensing cost as much (or more) than vSphere Standard. At that point if you don't need/want Redhat support VMware looks a lot more attractive. Oddly the only STIG for Suse I'm seeing is for Z series.
