You are 100% correct in your assessment, the situation is exactly as you describe it!
I don’t think I ever encountered the second problem you describe, the non-scheduled scrubs of non-boot pools, not too sure why, but your whole py-libzfs reasoning makes perfect sense to me. In any case, I only experience the bug for the boot pool, and applying the patch manually fixed it.
For the rest, yes, it’s an absolutely sad state of affairs for TrueNAS CORE, such a fine product on top of such a fine OS!
But, you know, what do I know? I’m just a user (of a TrueNAS CORE rig that has been running workloads for several SMB shares, an ever growing number of jails, a few VMs, one of them even with PCI passthrough for three NICs --my pfSense router–, and a few other things, and it’s been so incredibly rock solid ands stable for almost 10 years now, that frankly at times it even gets quite boring because it barely gives me any work!).
As I’ve said many times before, I’d love to participate in this project, even in its early efforts, I’m a web & devops & platform dev very comfortable with FreeBSD, so I’ll surely be able to contribute some, time commitments notwithstanding of course, but my offer has gone ignored so far, unfortunately.
Sorry for saying this, but messaging CORE “was not proceeding into the future” is by very far NOT what iX did, at all!
In my opinion it’s very hard to not see the treatment that CORE received as the classical bait-and-switch that unfortunately has become so standard in this industry, because at first there were words all over the place from iX about how CORE was not going to be abandoned, ever, and describing SCALE as no more than just a parallel product for a different audience, with assurances that both would evolve into the future together, in parallel.
A few of us very vocally said that’s not what it felt like, pointing at tangible practices coming from iX like development effort, Jira activity, etc., etc., kindly requesting clarity from iX, and all around we were told we were overreacting, even called paranoid at times… and, wha’da’ya’know, fast-forward just a few “paltry”, to use your own words, years, and guess what happened…
But that’s the problem, and that is, I think, fundamentally what has people upset. You[1] have not clearly messaged this. To be as charitable as possible, you’ve been evasive[2] about this ever since you announced SCALE.[3] Even in the announcement thread for Fangtooth CE, you pushed back on my interpretation of that announcement as “CORE is dead.” But it is, even more so than the Norwegian Blue Parrot.
I’m loath to speak for anyone but myself, but I think I can in this case safely speak for a good fraction of the community in saying that that you’ve severely damaged what trust we had in you,[4] and your continued insistence on pretending that your messaging has been clear all along just isn’t helping. And yes, this is rehashing old discussions, but the pretense continues.
plural, that is, iX as a whole, not necessarily you personally ↩︎
and I think a more accurate assessment would be “dishonest” ↩︎
“We didn’t know how successful SCALE would be,” you say. Even if that were true five years ago, it became evident in a hurry. ↩︎
though, speaking for myself, there have been several contributing factors to that result ↩︎
Since we’ve veered so highly off topic here, I’ll wrap up this thread with the following. Again, we were well aware that some of our more BSD die-hards would find this hard to accept in the long run. (Hence my joke about there being an echo in here) Some folks care more about the particular technology under the hood than they do the functionality being offered. I get it, I like to nerd out on things as well. But what we found along the way was that the number of folks who fall into that camp is a very, very small minority (but vocal here on the forums). I think the rampant success of the SCALE and now CE adoption has convinced us of this fact, that it’s features and stability that drive usage and adoption. Not BSD vs Linux, or Systemd vs rc.d or whatever backend implementation details we on the nerd side like to argue about in our spare time
