On Wed, Jan 10, 2007 at 10:39:33AM -0500, Clark, Patricia wrote:
> Now you mention that your upgrade included an architecture change. When
> is Solaris not the same? When you change architectures. Sparc and x86
> platforms are not interchangeable.
>
> No, you cannot Upgrade when changing architectures. All of your indexes
> will need to be manually recreated with scanner, et al. This has been
> discussed at length in this list over the last year.
I'm not sure that's entirely true, Patti. In order to migrate
cross-platform, you have to do it as disaster recover procedures rather
than straight copies/tars of files. (The process of getting the data
on/off the tapes mediates the platform/endianness, at least as I
understand it.) You can move the indexes, but only with nsrck -L7 (or
scanner, which is the longer way about it.) I may not have followed the
thread that closely, but if you're only copying files around, it's not
going to work well. (On the flip side, if you say in the same
architecture, a simple shutdown, copy, startup should work.)
When we moved from Solaris (Sparc) to Linux, I was able to successfully
do nsrck restores for some indexes. I just didn't bother for most
systems unless they needed a restore and let the indexes build with new
data over time. I recall from the experience that I needed the
/nsr/debug/no_striped_recover for it to work (or maybe it was some other
post-transition restore issue.) I still don't know why I need that, or
what it's doing, or what it's stopping me from doing, but "it works."
Also, in some of my dry run tests of the mmrecov on the new system
(prior to our transition, when I was trying different things,) keeping
the version the same was important. Even a small point release (I don't
remember if it was 7.11 to 7.12 or something like that) botched things
up in the DR restore. I would absolutely do them as two different
steps; either upgrade, then migrate or migrate then upgrade but keep the
versions consistent through the migration.
>From 6.x to 7.3, I'm not sure which I'd recommend doing first. Probably
the transition, make sure things are working, and then upgrade. Prior
to the op (too late?), I'd make an exhaustive list of things to test and
make sure they're all working before moving to the next step. (Can you
backup, can you restore, can you restore from a pre-migrate/upgrade
backup, can you restore indexes from a pre/post migrate update, etc.)
Dave
To sign off this list, send email to listserv AT listserv.temple DOT edu and
type "signoff networker" in the body of the email. Please write to
networker-request AT listserv.temple DOT edu if you have any problems with this
list. You can access the archives at
http://listserv.temple.edu/archives/networker.html or
via RSS at http://listserv.temple.edu/cgi-bin/wa?RSS&L=NETWORKER
|
