It doesn’t? What system do you think Hex will go with and will that system will be better/easier/more “click-click”/ longer lived/ etc. than what SCALE users are doing if they’re “only going after the home-user”? Will iX adopt Hex’s “system” if they have something better (and visa versa). I suspect they are hoping/betting on you to supply them. What friction will this cause?
A soft-fork which just replaces the GUI head and adds “recommended defaults” during build (sort of like how iX does their products; they pull from a base and add/remove from there).
A hard-fork which everyone knows.
How will they get around current limitations/shortcomings in iX’s API especially when it comes to apps and wizards and will it be open/two way?
Everyone’s guessing at this point because there are so many factors no one can know for sure without being in the loop. I’m also thinking about bigger picture/longer term questions.
Why would you say it’s overly celebrated? It solves a very real-world problem, it just so happens to be a problem that’s more likely to impact a home user than an Enterprise.
I don’t think I agree. It isn’t perfect, to be sure–it’d be much nicer if it actually re-wrote the data at the new data/parity ratio–but it answers a need that a lot of people have had[1]. Related needs not yet addressed are drive removal and conversion to a different vdev type (mirror to RAIDZ, RAIDZ1 to RAIDZ2, etc.).
Sure, many of these could be avoided or minimized by better pool design. But they remain things that smaller-scale users find helpful (or even essential), that some other filesystems promise[2], and ZFS just doesn’t do.
Not to mention the very large middle ground between “home user” and “has their own data center with racks full of servers.”
one which, in earlier years, had lots of people messing up their pools by striping in a single disk to a RAIDZ pool; more recent versions of Free/TrueNAS have added “seat belts” making it harder to do this ↩︎
Yup, I think you’ve made some good points about what’s still “Missing” in ZFS for a less-than-enterprise use-case. I’d like to think that a player like HexOS would help foster development of those missing features.
That’s the main reason I am excited for HexOS, not because I have any plans to use it directly, but simply because more mainstream adoption of ZFS will benefit the entire ZFS ecosystem.
Winner! Not a fork, its even using the exact same TrueNAS SCALE images we are releasing. We will have more to say in that in the future, but we’re well aware of the HexOS news and have been collaborating with them for a while now. It’s aimed at an audience which is super non-technical, I.E. if you already use TrueNAS today it’s probably not for you.
HexOS will have the TrueNAS full system installed. It’s just a click away if somebody wants start using the TrueNAS UI directly for more advanced features and functionality.
In a product matrix in online shops regularly the most powerful model of product X is placed in the rightmost column. And when you configure all options to the max, that’s frequently at the bottom.
So “bottom right”: most powerful, most capacity, most expensive.