I was talking about iSCSI support, not how the ZFS themselves pools are built. Yes, ZFS *does* prefer raw disk backing (due to certain management simplifications), but it does not need it. In fact, it's quite possible and people do frequently run ZFS on top of RAID arrays.

I've confirmed it by actually reading the source code. Even a quick look at the manpage will tell you the same:

Keep perspective. Are you really going to build a box like that with just one 2 GHz quad-core CPU?
I have pushed 4GB/s through a SAS SSD array on ZFS, but even so I maxed out on other stuff way before the CPU and much less checksumming ever began to be an issue (e.g. had to go through two LSI SAS 9200-8e HBAs, because one maxes out the PCI-e 2.0 x8 lanes; with two HBAs I maxed out on the two 6G SAS links to my JBOD). That the point of my post. I've yet to see a system which is constrained by the checksumming in any meaningful way.

Sun still had closed bits in the OS kernel when OpenSolaris was CDDL'd, so GPL was a non-starter because of GPL's "infectious" nature of spreading to all source that makes up a project. CDDL is more permissive in this - it simply forces you to keep the free bits that you got free, but doesn't expand to anything outside of the source files you got.

iSCSI doesn't need to be baked into ZFS, in fact, even on Illumos it isn't. It's in a completely different subsystem and will happily work with any block device as its backend storage (be it a physical drive, a ZFS zvol, a loopback block device or anything else, really).

You can thank your package maintainer for this. ZFS never ever ever upgrades the on-disk format silently. You always have to do a manual "zpool upgrade" to do it. It'll tell you when a pool's format is out of date in "zpool status", but it'll never do the upgrade by itself.

Again, this is not ZFS' fault, it's your package maintainer for auto-upgrading all your imported zpools. ZFS never does this by itself.

The overhead is barely there at all. I've measured the performance of the default fletcher4 checksum on a modest 2GHz Core 2 CPU and it comes to around 4GB/s/core. Now given that most CPUs now come with 4 or more cores, in order to get the checksum to be 10% of CPU overhead, you'd have to do be doing around 1.2GB/s of I/O. Needless to say, you're not ever going to get that even for fairly high-performance boxes.

What answer? You gave a few links to wikipedia with zero analysis to support it. You claim shit like "obviously we can get some energy back from the steam turbines too" without actually knowing and showing that doing so is feasible.
Look, it's really simple. Give your best case analysis for what you propose to do. No links to wikipedia and vague statements like "seems your main technical objection is being worked out with a 780 KW beam for spallation". Tell me exactly how you propose this to be done and how that adds up quantitatively. How much power, in what accelerators, how much would such a system cost and how would you propose we construct it. No more evading. Talk to the point.

Because you're being so disingenuous. Any time I show you how wrong you are using very basic mathematics, you just change the subject, or make a new equally outrageously wrong claim.
Anyhow, now that you're finally making claims which are at least vaguely quantitative and testable, we're at least getting somewhere.

First off, I hope you meant beam current, not luminosity (since that's a property of accelerators with detectors), since that's the property that actually tells you the number of particles in the beam and therefore how many nuclei you can affect. Second, the rate at which you do it doesn't change the total energy investment needed, it'll still cost about $20bn/ton. Oh and how much such a large number of facilities would cost to build and operate is a whole other matter. Oh and the transmutation products might still very well produce a significant amount of decay heat, so you'll have averted exactly zero risks of meltdown (though this depends on the details of your proposal).
But go ahead, present your detailed numerical analysis. Perhaps you have some amazing physical insight that makes this all wonderfully efficient, safe and sensible.

Shut up it is then.

The MEGAPIE accelerator you linked gives tops ~1mA of current and substituting tritium for protons lowers that by about a factor 3. I'll spare you the numbers, but in effect, to consume 1 ton of long-lived fission products this way would take on the order of 50000 years to the tune of some $20 billion per ton just for the power needed to run the system. If you think this is even remotely practical, you're an idiot. And this is the absolute best you can do, ignoring all practical issues of handling large quantities of radioactive tritium and fission products, chemical SNF separation, Tritium's limited half-life (so the need to regenerate it), cooling requirements, physical arrangement, etc.
The reason MEGAPIE was built and your crazy fission-product incinerator wasn't is because the guys at CERN are actual scientists and you're just an Internet armchair expert.

WHAT THE FUCK. I have no understanding? When it is you who can't show the first thing about anything quantitatively and just deflects from the topic? You're like a textbook example of the Dunning-Kruger effect. You're so overconfident in your statements, yet when pressed, can't support any of them with hard data. All you can do is google stuff you vaguely half understand, ignoring all practical problems with it, forgo any and all mathematical analysis and declare victory.
I'm done. If you post your mathematical analysis of such an incinerator system, I'll respond. Anything else, I'll ignore.

I didn't agree to look up numbers, I wanted you to support your claims with actual calculations, even if only in the ballpark region, by letting you present your best case. This isn't a 20-questions type of thing where you interrogate me on what I think is workable. It's you laying out your case and presenting factual data to support it and since this is hard science, it better be quantitative. More and more I'm beginning to think that you can't do it, so you're just running your mouth, diverting attention and changing the topic. I this paragraph written up showing you how using the MEGAPIE accelerator to do the task you propose (radiocesium destruction by tritium acceleration) was silly on its face and how even trivial calculations show that it's just not practical, but I'm not going to do your work for you.
We have this saying in science: put up or shut up. So which is it gonna be?

Please stay on topic. I'm beginning to think you actually can't produce the numbers to support your claims.

1. I wasn't the one that calculated that
2. If the speaker disagrees with the assumptions, challenge the assumptions rather than resorting to an argument from authority.
Must can be wrong and has been wrong. Most likely though we're not getting the full scoop in the story here. There's probably a load of caveats and detail to their plan that cannot simply be boiled down to a simple marketing snippet. Unfortunately, that's how marketing works.

So what are the numbers on that?

So what are the numbers on that?