I agree that we should log this at a higher level that isn't visible by
default. What do you think Martin?
Tom
On 10/15/2012 9:01 AM, Hong Zhang wrote:
> Hi, Tang/Jeremy
> I was able to reproduce this, but when I tried to execute deploy command
> the second time (after the application is undeployed), I don't see this
> message anymore. I suspected this has something to do with the command
> caching feature (which is intended to optimize command execution
> performance). There was a new deployment command option added sometime
> last week, so the command caching detected that change and started to
> use/cache the new command option signature. I assume the deploy command
> executed successfully despite the message?
>
> Adding Tom/Martin to the thread to see if they agree. Probably we should
> not show the message ("Command signature of deploy command was changed.
> Re executing with new metadata.") to the user as it could cause
> potential confusion and user does not really need to know what happens
> behind the scene for the command caching.
>
> Thanks,
>
> - Hong
>
> On 10/15/2012 8:37 AM, Tang Yong wrote:
>> Hi Jeremy,
>>
>> Thanks your confirmation and I want to ask whether dev tests contained
>> such a singleton ejb's deployment case or not?
>>
>> From your confirmation, it seemed to have a regression on deployment.
>>
>> Thanks!
>> --Tang
>>
>> lvsongping wrote:
>>> Dear Tang：
>>>
>>> I have tried to reproduce it and found it can be reproduced in the latest
>>> version of GFV4.
>>>
>>> I also tried to reproduce this phenomenon in the gf trunk before October 9,
>>> 2012 and found
>>> it can't be reproduced. Anyway, I also tried to deploy any other kind of ejb
>>> samples, such as
>>> stateless ejb sample. The similar error messages can't be thrown out.
>>>
>>> According to the phenomenon I have referred above, I think it may be caused
>>> by any changes
>>> about the trunk between October 12, 2012 and October 15, 2012 and it could
>>> be a bug about
>>> deployment.
>>>
>>> Hong:
>>> What's your opinion about this phenomenon?
>>>
>>> Best regards
>>> -Jeremy
>>>
>>> -----Original Message-----
>>> From: Tang Yong [mailto:tangyong_at_cn.fujitsu.com]
>>> Sent: Monday, October 15, 2012 8:53 PM
>>> To: Hong Zhang; dev_at_glassfish.java.net
>>> Cc: Sahoo
>>> Subject: About singleton ejb's deployment
>>>
>>> Hi Hong
>>> CC: Sahoo,
>>>
>>> While investigating GLASSFISH-19152, there is a problem about singleton
>>> ejb's deployment and want to discuss with you and deployment team.
>>>
>>> While I execute the following command using the attachment(14/Oct/12 's
>>> version) of GLASSFISH-19152, the following result happened.
>>>
>>> asadmin deploy D:\singleton-osgi\target\singleton-osgi-1.0.jar
>>> Command signature of deploy command was changed. Re executing with new
>>> metadata.
>>> Application deployed with name singleton-osgi-1.0.Command deploy executed
>>> successfully.
>>>
>>> asadmin undeploy singleton-osgi-1.0
>>> Command signature of undeploy command was changed. Re executing with new
>>> metadata.
>>> Command undeploy executed successfully.
>>>
>>> I felt suprised to "Command signature of deploy command was changed. Re
>>> executing with new metadata.", so I want to ask you whether deployment
>>> command usage has a change or not?
>>>
>>> BTW: I used my building snapshot based today's gf trunk.
>>>
>>> Thanks!
>>> --Tang
>>>
>>>
>>>