[rescue] Potential SVM issue in Solaris 10; anybody got a spud?

[velociraptor]

On 6/5/06, Skeezics Boondoggle <skeezics at q7.com> wrote:
> well, in general, yeah... but how much are you _writing_ to your / or
> /etc?  turning off journaling for partitions that are largely readonly
> shouldn't be much of a loss... but then, i'm still used to the pre-bloat
> solaris 7 era, where i could mirror a pair of 9gb drives, create a 1gb /,
> a roomy swap and /var, and still have room left over for a local copy of
> /usr/local if i wanted to (rather than nfs mounting that).  i suppose if
> folks are doing the "modern" thing and creating a swap slice and lumping
> the rest into /, then yeah, you'll probably want to journal it.

I am of two minds about it--I'd partition out /var if logs might
become an important issue on the box, otherwise, just / if I have big
disks.

> i've been stuck in suse hell at my new job, and the last _significant_
> solaris work i got to do (for money, anyway :-) was solaris 7.  i've been
> slogging through solaris 8/9/10 and trying to come up with a clean,
> minimal, efficient installation - fully characterized by jumpstart
> profiles and cfengine rules that cut out all the fat i don't need... it's
> slow going.  i'm still not convinced i shouldn't go ahead and update all
> of my solaris rpms that were so carefully and _consistently_ built for
> solaris 7; everyone says "just use the companion cd" or "one of the
> pre-built freeware sites" to save time... but none of them get it right.
> still.  /opt/wtf?  /usr/sfw?  are you fscking kidding me?  puke.

I suggest you investigate Flash Archives.  Jumpstart + profiles is old
and slow comparatively.  You still use it for netbooting and
initiating the flar install, but flar'ing is ever so much faster.

As to where the software ends up, does it really matter?  I mean, I
know we're all so tradition bound by "/usr/local", but you know,
things change.  Not to mention the fact that you *can* change
locations with blastwave, it just takes some work, whereas SFW is
almost always in /usr/local.

> y'know, i remember doing this in the past but not having it be an issue...
> it was auspex-labeled seagate disks (520-byte formatted!?) that i updated
> and reformatted first on a sun, then later moved off to a netapp which
> couldn't use the auspex disks directly (possibly because i was running an
> old ontap release that didn't yet support 520-byte "bcs" disks?) but
> _could_ after my intermediate steps.  of course, upon detection, the
> netapp immediately put its _own_ new firmware on each drive!  i've used
> netapp disks on a sun before with no problems... got an old filer around?
> it has the advantage of letting you run disk_fw_update against a whole
> shelf of drives at once! :-)

The chances of landing a job where I support NetApps seems small these
days, more's the pity.

> as i recall, the particular drives in question had a sun-provided firmware
> updater available; almost all of their disks have an associated patch that
> can download sun's latest recommended firmware.  (sadly, those _might_ be
> in the contract-only section?)  i don't recall that step being difficult
> or remarkable... so even if the inquiry returns "COMPAQblah", if the
> underlying "STblah" (or ibm, fujitsu, whatever) disk is supported, i think
> the firmware patch will usually find it.
>
> but it's been a while, and i could be talking out of my... hat. ;-)

Welp, I'll have to take a poke around sunsolve.  I don't recall seeing
anything like that, but I've never really looked for this particular
kind of beastie.

=Nadine=