Illumos to drop SPARC Support • Hacker News Board

Re: Illumos to drop SPARC Support

#181

Post by ptribble » Sat, May 08, 2021, 9:23 AM UTC

yjftsjthsd-h wrote: post #139

ptribble wrote: post #138

Earlier quoted context omitted.

Tribblix SPARC will continue. While you lose any new features that go into illumos upstream, at least you don't keep getting broken by changes going into illumos upstream. (This isn't a commitment for all time, naturally. At some point the SPARC hardware I have will stop working. But it turns out to be solidly built and impressively reliable.)

So for the foreseeable future you intend to just freeze SPARC Tribblix on the last Illumos to support SPARC? I suppose it should stay ABI compatible with other components ~forever, so that shouldn't even hold back other pieces of the system.

Pretty much it, yeah. If there's anything useful upstream then it will get pulled in, and there's a still a lot of worthwhile cleanup of SPARC, but the kernel will be essentially frozen.

Re: Illumos to drop SPARC Support

#182

Post by nix23 » Sat, May 08, 2021, 9:47 AM UTC

arid_eden wrote: post #132

nix23 wrote: post #82

Earlier quoted context omitted.

>Torvalds even said so himself. Torvald's comment about ZFS was as uninformed as it gets...and he calls himself an FS-Guy ;(

His comment was more on the wisdom (or otherwise) of running an out of tree filesystem. I think its hard to disagree with him. He went on to say you would never be able to merge the ZFS tree with Linux. Again he's the one who would know what code gets in Linux. His only actual comment against ZFS was that benchmarks didn't look great - which is unsurprising given all the extra work ZFS is doing wrt data integrity tha…

Not everything needs to be in the Linux Kernel (and honestly i don't care if it is), looking at the past "linux-sound-system-tragedy" i would say, to make something outside linux is often much better (not 1000's different peoples who thinks it's better the other way around, and you are full of sh* anyway).

>His only actual comment against ZFS was that benchmarks didn't look great

What benchmark? Mines are looking pretty good, with a preheated Arc and with L2 especially...actually much better than any HW-Raid. Compared to Linus, there are Institutes with a bit more than a single 3GB git repository, and the crazy stuff...they need verified backups.

https://computing.llnl.gov/projects/zfs-lustre

If he ever has to improve a Lustre Filesystem of 55 petabyte with ZFS, he can come back, otherwise Linus...shut-up* and be happy with your ext4 (nothing against that one).

* A homage to the old linus-style of having a discussion.

Re: Illumos to drop SPARC Support

#183

Post by reg_windows » Sun, May 09, 2021, 1:50 AM UTC

corty wrote: post #34

qwerty456127 wrote: post #21

Sound crazy. Like if Windows had dropped x86.

Yes, but Sparc was on life-support as soon as Oracle bought Sun, and dead soon after. Sparc was already hideously expensive and slow compared to x86 during the Sun days. There innovation in Sparc came to a halt soon after T1, which also dropped most of the really nice features like CPU and RAM hotswapping. So you just got something expensive, slow and incompatible for five to six times the price. Oracle then proceede…

This is so very far from the truth. The 2000's were a poor decade for Sun's management of SPARC. They wasted hundreds of millions on the Rock chip which is a story waiting to be written someday (Marc Tremblay, the Rock architect, was being portrayed with god-like praise by Sun executives, always citing the mountains of patents he had but as we know, patents aren't a leading indicator of technical or business success). Then they foolishly latched the SPARC program to a Stanford professor's idea of Chip Multi Threading which had some good ideas but was not performant at the very time when X64 clock rates were rocketing, due to Intel and AMD's fab processes. I'm not an Oracle apologist by any means but Oracle revived SPARC R&D and made the chip competitive again. Larry even funded a lower cost S7 SPARC chip intended for high volume cloud deployments. And as has been mentioned, Oracle also supported (again, with millions of dollars) the SPARC software-in-silicon program, which provided hardware support for encryption, bad pointer detection, and tabular memory compression/searching (DAX). They tried but it was too late for SPARC. But it should never be said that Oracle didn't pump a lot money into SPARC after the Sun aquisition.

Re: Illumos to drop SPARC Support

#184

Post by ahl » Mon, May 10, 2021, 9:28 PM UTC

simcop2387 wrote: post #131

ahl wrote: post #95

Earlier quoted context omitted.

What is the latest on RAID-5/6 support for BTRFS. RAID-Z has its issues (saying this as the triple- and double-parity author), but it's been stable.

As far as Ive heard it's still pretty iffy. Since you were involved with the zfs side of things what are your thoughts on the upcoming zfs draid bits? I don't have specific need for them myself but they look really attractive for building a new pool to replace my aging drives.

Unfortunately I've been completely out of the loop on the draid stuff.