Anyone else notice that a lot more people are delving into FreeNas on ESXi (even including me to some extent)? I know it is not precisely a "new" trend, but there seems to be a large influx of threads regarding it...
Most of the regulars obviously know about the other threads and warnings. But just seeing what everyone's take is on being more "supportive" towards it?
The problem's the same as it was years ago when I wrote the anti-virtualization warning sticky. In general there's nothing wrong with a sufficiently clued individual virtualizing FreeNAS, but that probably includes needing some expertise with both FreeNAS and ESXi, and following a reasonable formula, rather than just pulling some random garbage out of your butt and making it up as you go along.
I didn't write the anti-virtualization stickies because I thought it couldn't be done... most of the FreeNAS instances here are virtualized and had been for quite some time by the time I wrote the warning.
The problem is that virtualization is every bit as twitchy and demanding as FreeNAS is. Sure, you can run FreeNAS on your repurposed old laptop and get it to share out some files, but do you really think that's going to be a good idea long term? I was seeing all sorts of people come in with the virtualization equivalent of that janky setup and then saying "but my hardware doesn't support VT-d, and RDM was working fine... but my disks no longer seem to have data on them, bad FreeNAS, and how dare you tell me the data might not be recoverable, those were my kids baby pix on there."
How many people do we see coming here with poor hardware selections for FreeNAS? Failing to set up things like SMART? Not burning in their systems? Not being willing to do the research to do things right? That's right... lots of them.
Virtualization adds another layer of butthurt on top of that, because the guys like
@joeschmuck who actually have the patience and discipline to work through things one step at a time, carefully researching and learning along the way, are in the minority.
The people who come to this forum, quietly read the virtualization warning, then go off and read the linked articles on how to virtualize, and then have success with that, there's a lot of them... and we never hear from them, because it just works. When you're doing it right, it's all good. The ones I'm afraid of are the ones who come into the forums, can't be arsed to read the resources that are available, or worse, don't *understand* the resources that are available, and then start asking questions. There is a point at which you probably just shouldn't try.
If I sound bitter and cynical, it's because I've been through too many cycles of explaining to people why their half-arsed setups won't work right, then having people try to do it anyways and continue to explain to me how I'm wrong despite things obviously not working right for them, as evidenced by continued questions. Or, worse, the data losses of those who liked to argue RDM was the way to go, until they lost their data and it became unrecoverable.
The simple fact is that virtualization offers the opportunity for lots of additional complexity, which is in turn room for lots of additional things to go wrong, especially when you have people who are doing "creative" things to get a little home lab box to "work right." "I can't afford more memory." "My box doesn't support VT-d." "I don't have room for a PCIe card." etc.
The people who know ESXi and are sufficiently versed in it to look at my virtualization stickies and to understand all the things I said? They have no problems virtualizing. But they also typically don't participate heavily in the forum. And the ones who do, hey, I love talking about the ins and outs of this stuff with them.
