KVM or VMWare
-
@rjt said in KVM or VMWare:
@francesco-provino Amazon Web Services may have a slight disagreement with you on whether KVM or XEN is suitable for business.
KVM and XEN are suitable for the business case of an hyperscaler of course, but the question of @WLS-ITGuy was literally "We're getting ready for our server refresh and along with that our license is up for renewal for VMWare. I am curious to the benefits of KVM over VMWare." -> so they are a small shop already using VMware.
It totally makes no sense to switch from VMware to KVM or Xen-based solutions in his business case.
-
@rjt said in KVM or VMWare:
@francesco-provino Amazon Web Services may have a slight disagreement with you on whether KVM or XEN is suitable for business.
As do every major environment. When you get big, or small, VMware makes little sense. It borders on the absurd. But in the middle tier, huge companies (not SMB) that aren't yet the massive scales of AWS, Google, or the big Wall St. banks, VMware tends to play nice because they have skills and value to automation, but can't write their own solutions. That's VMware's core market. Get smaller than where automation makes sense, which is 95% of businesses, and VMware is in the way of efficient operations instead of aiding it.
The biggest problem is seeing IT as a checkbox, a one size fits all where we just choose a vendor to sell (whether we are paid directly or not) and don't ask about the customer size, needs, use case, workload, etc. and see everything as "this one approach will always work" when, as IT, the one clear "always our job" is to evaluate that need and choose the solution accordingly.
-
@francesco-provino said in KVM or VMWare:
@rjt said in KVM or VMWare:
@francesco-provino Amazon Web Services may have a slight disagreement with you on whether KVM or XEN is suitable for business.
KVM and XEN are suitable for the business case of an hyperscaler of course, but the question of @WLS-ITGuy was literally "We're getting ready for our server refresh and along with that our license is up for renewal for VMWare. I am curious to the benefits of KVM over VMWare." -> so they are a small shop already using VMware.
It totally makes no sense to switch from VMware to KVM or Xen-based solutions in his business case.
I'd say the exact opposite. Now, that he already has VMware gives VMware an edge, where it would never have made sense to put in VMware in the first place, but since it is already there the least effort is continuing with it. But the effort to switch is half in the evaluation, and half in the doing. We move customers off of VMware to KVM regularly and it is fast, easy, and once done, it reduces their risk and cost (mostly by reducing support, but also reducing the need for third party software and, obviously, VMware licensing itself and the biggest cost, consulting hours for VMware licensing.)
In the OP's scenario, VMware is a solid consideration. But does KVM have absolutely crystal clear advantages? Yes 100%. At the OP's scale, VMware is technical debt. The question is only... is the debt too great to bother eliminating? Do they live with a "good enough" solution, or invest in a longer term, easier to support, lower cost, lower risk alternative? The problem is that it's only a little less support, only a little less cost, only a little less risk. So it is a hard comparison.
But the one thing we can guarantee is that no solution is an obvious choice. Not VMware, not KVM. It's a business decision based on a lot of small factors.
The biggest single reason to avoid VMware is the ecosystem around it that is so unhealthy and pushes it and other paid for, licensed, add ons and costly support models very, very strongly without generally any evaluation of the customer's needs. Because VMware is a product sold through the channel, and one with massive profits for the sellers and supporters of it, it creates a system of people and vendors industry wide who push it for their own interests and that's dangerous to customers.
-
@olivier said in KVM or VMWare:
@Francesco-Provino by scale I meant the scale of Xen core developers (ie headcount). It's not that much especially compared to Xen adoption at such… scale
So if you did a lot of things with a relatively small team, this is a pretty nice clue on what could be done with more focus and people! (and yes, Citrix was completely unfocused on Xen, only few years after acquiring it).
But this is clearly changing (the situation, not Citrix
).Something that IT people often don't realize about software engineering is that often the most work, and the best work, is done by small teams. The Mythical Man Month is rarely read in IT circles, only development ones, and the idea that a two man team can easily outpace a 100 man team doesn't seem logical unless you've worked as a developer.
I "recently" had this experience with a customer who trusted a small team, but couldn't understand how a small two man team starting from scratch could compete with a well funded, twenty year old, 65 man team making head to head products. But the two man team had caught up in six months, which was expected on their end.
Partially not dealing with technical debt, and partially by avoiding the network communications effect, and using only 10x people instead of lots of 1x people... all things developers know well but outside of software engineering is almost totally unknown.
VMware actually struggles with how to remain competitive while being so large, rather than the other way around. I'm not saying that they don't overcome it and do a great job. Only that it is a challenge for them that smaller teams don't have. WHen you make software, getting big, makes things hard if you want to keep moving forward. You can do it, but it isn't a "get big and things get easy" situation like in most other industries.
-
@stacksofplates said in KVM or VMWare:
@pete-s said in KVM or VMWare:
It isn't the ability to automate that is the problem. It's the availablility of easy to use tools that is the problem.
Thats the whole point I'm making.
KVM is hard to automate. Not that it's impossible, but the tooling doesn't exist to where you can easily automate like with VMware.
Agreed, and I don't think that that's the point of concern here. The issue at hand should be "does that automation that VMware offers get used by or should be used by the OP?" I believe that the answer is no to being used today and likely no to should it be used. It's a very small deployment. The overhead to the automation, even when you have VMware, is too high. And regardless, even if we agree that it should be used, probably because an MSP/ITSP is brought in to effectively make the environment larger and changing some of the scale discussions, the bigger question would be "will the OP's environment opt to do that anyway?" If that answer is "no", in the practical sense, then the automation point becomes moot.
I "think" we can all agree that VMware has better standard built in automation. And that KVM is completely automatable if you put in the extra, non-standard effort. So if we were considering standard automation then VMware would have an important edge in that area. That point shouldn't be in dispute. We can argue how close KVM gets, while still being behind, sure.
But the key point here, for me, is that I believe based on knowing the environment a bit that that automation is not, and won't be, used if VMware remains.
