On 09/18/09 02:23 PM, [email protected] wrote:
11221 beadm create core dumps on OpenSolaris 122 SPARC
http://defect.opensolaris.org/bz/show_bug.cgi?id=11221
When using build 122, an attempt to create a new boot
environment (BE) via beadm(1M) or indirectly through "pkg
image-update" may result in a "segmentation fault".
On 09/20/09 05:48 PM, I wrote:
No segfault - get this instead with build 122 (sun4u):
# beadm create test
Unable to create test.
Mount failed.
After much pain, managed to install snv123. Now beadm /does/
segfault when creating a new BE (same pstack). I see the bug name
has been changed now and it won't be fixed until next year :-(
Is there any reason why beadm shouldn't be somewhat independent
of release much as pkg is so that it can be fixed?
The odd thing is that despite the segfault, it seems to create
a perfectly good BE. Does this mean the segfault can be simply
ignored, or is something nasty going to happen under the covers?
I guess it really means that an upgrade from snv123 to snv124
will be impossible since it won't get past beadm, and none of
my previous BEs can read the zfs rpool because I didn't think
to do zpool create -o version=xx rpool <disk>.
This is one of the most infuriating and time consuming bugs I
have ever encountered in many years of using Solaris. I hope
it gets fixed in such a way that it isn't necessary to fall
back to an ancient release to do an update...
Sorry about venting - I know you are all doing a great job, but
this one has been so frustrating. Perhaps the best workaround
would be to install snv121, but it isn't obvious how to do that.
Cheers -- Frank
_______________________________________________
indiana-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/indiana-discuss
