Thank you, Aaron.
There are multitudes fleeing VMWare right now. Many have only heard of Proxmox and not used it. I imagine Proxmox is interested in embracing them with open arms. And yes, I have several clients that will be purchasing commercial support for whatever replaces VMWare for them. I've been using Proxmox for years and I'm happy to talk about the benefits of it. It would be a shame if these users, who only have the experience of ESXi, are shot down in this way for asking really simplistic questions. I'm sure the paid support wouldn't treat them this way, but evening paying customers will visit the community forums.
Yeah, I'm angry right now, because as much as I like Proxmox I'm now nervous about recommending it to anyone if this is how the community will respond to even dumber questions than mine.
We get a lot of questions, that's for sure. Some of them on the level of how to setup ZFS or the most basic plug-in virtual networking. We help those people because they are actually trying to use PVE how it is meant to be used. Your question is closer to how can we make PVE be more like VMware. You aren't really trying to use PVE how it should be used, nor do you seem interested in how PVE goons pack and deploy their stuff. Your question is basically
I want support for a VMware-formatted export in the PVE GUI. Please acknowledge that you understand what a difference that is.
VMware will be getting what the market will bear from its well-trained audience which has a specifically and deliberately limited understanding of virtualization and can only work within its own paradigms. Broadcom knows better than anyone how true this is and how willing the public will be to pay for their comfort zone, especially after evaluating the market. Customers looking for VMware alternatives, especially free ones, will need to get just a little more flexible and realistic about how unproductive it is to want to cling to VMware's memory and their VMware skill set while at the same time trying to move to something a lot more agnostic.
Proxmox has never wished or attempted to poach VMware's customer base nor does it purport to be an effortless drop-in replacement for VMware... but all of a sudden we get posts every day, from VMware users, asking for almost exactly that. Maybe one day there will be a sticky post about VMware. Rarely do the VMware users stop to ask themselves
is there a better way to do XYZ or do I just want it to be done 'the VMware way' and to post on the official forum asking them to put VMware functionality into their free product.
Proxmox is totally free software. Your support subscription buys you access to the enterprise repo and some professional ticketed support. You can also buy Proxmox training, who knows with any luck the trainees could even develop an appreciation for PVE in contrast with VMware. You can ask for new GUI functionality all you want, but it's in no way tied to any funding brought in by support subscriptions, and the promise of a larger install base is really not the leverage you think it is when it comes to deciding what functionality to add. Again, be realistic.
If OVF was important, it would have been integrated eons ago. Outside the rapidly crumbling Potemkin Village of VMware, OVA/OVF is an entirely irrelevant format. Even Hyper-V does not readily work with it. No one besides Workstation and ESXi really even uses VMDK that much, except as MAYBE an intermediate format... Why hasn't Microsoft done anything about it? Why is there no interest? I don't know, but I can assure you the level of interest is quite low. It is little more than a zip file with a vmx and a vmdk. It would not serve the PVE community to keep it on life support for the sake of VMware refugees, a majority of whom do not represent any revenue to Proxmox.
Software vendors that distribute paid products via OVF are more likely to come up with more novel and more flexible methods of distribution, and in less time, than you are to see other platforms integrate OVF to the degree of, by your description "as simple as ESXi." I'm afraid the further expectation, if I'm not reading too much into things, of operating PVE successfully while never having to do any "CLI level work" is misplaced as well. Even VMware could not promise that.
Until then, parse the VMX and build a machine of equivalent specs. Convert the VMDK to what is appropriate for your storage, and go. 90% of those who are commonly understood to be of an intermediate skill level PVE admin would just do this and move on. The PVE vs VMware culture can almost be entirely boiled down to one group of individuals totally frozen without their 5-page wizards and the other group who can problem solve and just work through things.
If I were to suggest why this will never make it to the GUI....principally, there are far too many combinations of storages for PVE to support, and PVE has almost no hardware requirements at all, certainly not strict ones. VMDK is the least preferred format for virtual disks so we have to convert the template disk to what is appropriate for the destination storage. The OVF could be for a little firewall or TKL appliance or it could be a production file server. The OVF is arbitrarily and potentially impossibly large, and to ask that pveproxy staging the uploaded data from the web browser in /tmp, and it's not a good idea to assume /tmp is a good place to receive an OVF and converting a VMDK here. New users even have trouble importing ISOs this way rather than just using SSH/SFTP.
