Thanks, that fixed it!
One more odd thing I found, if I try to start the service (using
systemctl or targetctl) and it fails, the next time I try to start it,
my shell gets stuck. No amount of ctrl+c or kill -9 on the PID (from
another session) seems to fix this, I'm not even able to reboot
cleanly, I have to manually reset the server. I'm not sure if this is
directly related to fedora or not, but I removed the target-devel
mailing list as requested.
On Fri, Apr 3, 2015 at 8:27 PM, Andy Grover <agrover(a)redhat.com> wrote:
On 04/03/2015 03:28 PM, Dan Lane wrote:
>
> That took care of half the problem, but it's still failing because of:
>
> block/data: Cannot find attribute: fabric_max_sectors, skipped
>
> I tried adding it but either I'm putting it in the wrong location or
> my syntax is wrong, please advise.
Hmm? I think this means saveconfig.json contains fabric_max_sectors, but it
doesn't exist as an attribute for LIO. So I'd think you'd want to remove it
from saveconfig.json, not add it.
BTW since this is a -fb -specific issue, we can spare everyone else and use
the targetcli-fb list (CC'd) if further discussion is needed.
-- Andy