On Dec 3, 2007 10:28 PM, Lisa <zsh99 at yahoo.com> wrote:
> I was experimenting with ldom 1.0.1, ..after creating a test ldom, I tried to 
> enable the virtual console server but it refuses to start (see details 
> below). What am I missing?

My guess is that you didn't create a virtual console concentrator.  For example:

# ldm add-vcc port-range=5000-5100 primary-vcc0 primary

There are likely other useful hints at
http://blogs.sun.com/bradb/entry/ldoms_for_the_rusty.  I haven't read
this thoroughly, but at first blush it looks to have good stuff in it.

> # tail /var/svc/log/ldoms-vntsd:default.log
> [ Dec  3 22:10:56 Executing start method ("/lib/svc/method/svc-vntsd") ]
> [ Dec  3 22:10:56 Method "start" exited with status 1 ]
> [ Dec  3 22:10:56 Executing start method ("/lib/svc/method/svc-vntsd") ]
> [ Dec  3 22:10:57 Method "start" exited with status 1 ]
> [ Dec  3 22:44:13 Executing start method ("/lib/svc/method/svc-vntsd") ]
> [ Dec  3 22:44:13 Method "start" exited with status 1 ]
> [ Dec  3 22:44:13 Executing start method ("/lib/svc/method/svc-vntsd") ]
> [ Dec  3 22:44:13 Method "start" exited with status 1 ]
> [ Dec  3 22:44:14 Executing start method ("/lib/svc/method/svc-vntsd") ]
> [ Dec  3 22:44:14 Method "start" exited with status 1 ]
> #

Some error messages to stderr sure would be nice.  It takes way too
much guessing (or guided guessing with truss) to figure this stuff
out.

It looks to me as though this should have sent an error message to
/var/adm/messages.  Based upon the comments in the older of the two
revisions, there is/was aversity towards printing a useful error
message when exiting.  Look at the changes around line 363.

http://cvs.opensolaris.org/source/diff/onnv/onnv-gate/usr/src/cmd/vntsd/vntsd.c?r2=3151&r1=2748

Notice that the newer code tries to cause vntsd to signal SMF more
appopriately so as to not retry.  I'm not sure of the significance of
exit value 2.  svc.startd(1M) refers to $SMF_EXIT_ERR_CONFIG and
$SMF_EXIT_ERR_FATAL, which seem to only be defined in
/lib/svc/share/smf_include.sh as:

#
# smf(5) method and monitor exit status definitions
#   SMF_EXIT_ERR_OTHER, although not defined, encompasses all non-zero
#   exit status values.
#
SMF_EXIT_OK=0
SMF_EXIT_ERR_FATAL=95
SMF_EXIT_ERR_CONFIG=96
SMF_EXIT_MON_DEGRADE=97
SMF_EXIT_MON_OFFLINE=98
SMF_EXIT_ERR_NOSMF=99
SMF_EXIT_ERR_PERM=100

Seems as though it should exit with 96 rather than 2.  Of course it
should fprintf(stderr, ...) first.

-- 
Mike Gerdts
http://mgerdts.blogspot.com/

Reply via email to