Kedar,
All tests are passing, so I think we're safe. There is no guarantee
to any change.
I could insert code internally to admin service that enforces the
previous order if a flag is set to true (that's what
FORCE_FINISH_METHOD did). But I've never seen a problem in any testing.
Lloyd
On Jan 11, 2007, at 11:08 AM, kedar wrote:
> Lloyd,
>
> I also thought that you were going to ascertain that
> this change does make any difference to the semantics
> of how AdminService's various life-cycle methods are
> run. Did I hear that you needed some more time to do that
> or did you say it is guaranteed that there is no change
> of that sort?
>
> Kedar
>
> Lloyd L Chambers wrote:
>> Kedar,
>>
>> There were two minor changes we agreed upon:
>> - remove comment "it will call FeatureAvailability.registerFeature
>> () when ready."
>> - remove unused variable FORCE_FINISH_METHOD variable in AdminService
>>
>> I though our agreement was that I would take up the following
>> issues in subsequent work:
>>
>> (a) check into RunnableBase features vs features provided by
>> java.util.concurrent classes
>> (b) can ExecutorService take a name for the thread so that the
>> thread can be identified while running, blocked, etc?
>> (c) add a -D for random sleep feature in RunnableBase instead of
>> just setUseRandomSleep(), which requires a recompile.
>> (d) jmx.remote.whatever flag that auto-starts a connector--does it
>> work now?
>>
>> I've done (c) above. The others can wait as they are not
>> immediately relevant.
>>
>> Lloyd
>>
>>
>> On Jan 11, 2007, at 10:40 AM, kedar wrote:
>>
>>> Lloyd,
>>>
>>> I assume you've made the changes that we agreed upon
>>> during the review. This particular change is going
>>> to be followed by a review of RunnableBase by Concurrency
>>> Gurus (Brian et. al.) we have, right?
>>>
>>> Also, a better link to your background doc is:
>>> http://tinyurl.com/yyzlet
>>>
>>> Did you also get a chance to test it with -Dcom.sun.management
>>> properties for built-in JMX Connector?
>>>
>>> Kedar
>>>
>>> Lloyd L Chambers wrote:
>>>> Abhijit,
>>>>
>>>> I am doing a cvs update and then a fresh build and one final
>>>> quicklook-ee prior to committing my MBeanServer change. Kedar
>>>> has reviewed the key code.
>>>>
>>>> Abhijit, would you like to review the code yourself?
>>>>
>>>> QuickLook EE has been passing many times now, on more than one
>>>> machine.
>>>>
>>>> I'm planning on comitting this code today, probably by noon.
>>>>
>>>> Background:
>>>> http://www.glassfishwiki.org/gfwiki/attach/
>>>> GlassFishAdminReferences/SunoneInterceptorChanges.html
>>>> Bugs:
>>>> https://glassfish.dev.java.net/issues/show_bug.cgi?id=1409
>>>>
>>>>
>>>> Lloyd
>>>>
>>>> -------------------------------------------------------------------
>>>> --
>>>> To unsubscribe, e-mail: admin-unsubscribe_at_glassfish.dev.java.net
>>>> For additional commands, e-mail: admin-help_at_glassfish.dev.java.net
>>>>
>>>
>>> --------------------------------------------------------------------
>>> -
>>> To unsubscribe, e-mail: admin-unsubscribe_at_glassfish.dev.java.net
>>> For additional commands, e-mail: admin-help_at_glassfish.dev.java.net
>>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: admin-unsubscribe_at_glassfish.dev.java.net
>> For additional commands, e-mail: admin-help_at_glassfish.dev.java.net
>>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: admin-unsubscribe_at_glassfish.dev.java.net
> For additional commands, e-mail: admin-help_at_glassfish.dev.java.net
>