Alok,
Alok Aggarwal wrote:
> Hi William,
>
> On Fri, 13 Feb 2009, William Schumann wrote:
>
>> As Alok mentioned, the Orchestrator was ignoring the attribute
>> OM_ATTR_TIMEZONE_INFO.
>> http://cr.opensolaris.org/~wmsch/bug-4629/
>> http://defect.opensolaris.org/bz/show_bug.cgi?id=4629
>>
>> RE: GUI: The timezone is also being set in the GUI with a direct call
>> to om_set_time_zone(). I did not remove this, since setting it at
>> the earliest possible time is preferable. Removed setting of
>> OM_ATTR_TIMEZONE_INFO in Orchestrator attribute list so time zone is
>> not set twice.
>>
>> Added another fix - if the Orchestrator was failing in
>> om_perform_install() before the TI thread was started, AI would hang
>> forever.
>>
>> Tested AI and GUI, specifying valid, invalid, and missing timezone in
>> SC manifest.
>
> perform_slim_install.c: I think you want to additionally
> log what the timezone specified was and also word the
> error message (in case it wasn't) to be similar to the
> missing locale case.
It is logged by om_set_time_zone().
>
> Just to be sure, the GUI can actually set the timezone
> with this change?
Yes. The GUI calls om_set_time_zone() as before and the time zone is
actually set in testing.
William
>
> Looks fine otherwise.
>
> Alok
