[oi-dev] OI project reboot required

Jim Klimov jimklimov at cos.ru
Sun May 12 18:31:45 UTC 2013


On 2013-05-12 19:06, Garrett D'Amore wrote:
> So, out of curiosity -- *who* is actively running illumos on 32-bit kit today?  I'm not interested in hypothetical uses or kit that is sitting around in your garage waiting for you to do something with it…. I'm interested in people who would be immediately impacted and severely so if illumos were not available on 32-bit CPUs right now.  (To give a counter example: I have a 32-bit Atom netbook, that I have OpenIndiana on.   I turn it on once every year or so… if that often … so I can't seriously claim that I would be negatively impacted if illumos were to move to 64-bit only.)

Indeed, I can't vouch for such systems, even my old home-NAS which was
my first victim of illumos/OI endearment had a Pentium-D with x86_64
support (though likely not virtualization acceleration features -
which would IIRC require VMs to be 32-bit). This box would be in my
production today, had it not broken while I am on a prolonged trip
away from that home; but it won't be impacted by a 64-bit only OS,
indeed (it did have some VMs for a test farm though, and they might
be impacted).

Also I know of many small (SOHO) storage boxes which can be made to
work with OpenSolaris and illumos-based OSes, and of those only the
HP N36 and N40L have (known to me) 64-bit AMD CPUs and ECC support;
most other such boxes are built around Atom, and often 32-bit with
some 2GB RAM. While this is not "interesting" for intensive production
use, some users of these boxes as reliable (ZFS) home storage might
be hurt by move to 64-bit only OS. Arguably though, lack of ECC did
probably burn my home-NAS causing some corruptions poorly-explicable
otherwise. While I wouldn't recommend non-ECC ZFS NASes for any use,
at least as a newly built rig, people that have a black box which
just works, and aren't keen on spending time and money to buy and
setup regular upgrades, are kind of stuck with it until the HW dies.

My 2c of theoreticizing :)
//Jim Klimov





More information about the oi-dev mailing list