Re: upgrade testing

[Bryan Pendleton <bp...@amberpoint.com>]

Mayuresh Nirhali wrote:
> I believe, the upgrade model we follow implies that the upgrade is 
> possible *Only* from the previous major.minor release. So, the upgrade 
> test for 10.3 will only test any possible upgrades from 10.2.

That makes sense. Thanks for clarifying.

bryan

Re: upgrade testing

[Mayuresh Nirhali <Ma...@Sun.COM>]

I see. This explains why the other upgradeCases were not removed from 
the framework. So, no clean up is required.

then, I will file a JIRA to track supporting upgrade from any release 
for the current framework.

It will be worth considering converting this test to JUnit as a part of 
this effort!

mayuresh


Mike Matrigali wrote:
> I believe that the goal is to be able to upgrade from any release.
> I think what you are seeing is that tests haven't caught up with
> the upgrade support we want to provide.  I believe we should be
> able to hard upgrade a 10.1 db straight to a 10.3 db.
> 
> Daniel John Debrunner wrote:
> 
>> Bryan Pendleton wrote:
>>
>>> Mayuresh Nirhali wrote:
>>>
>>>> I believe, the upgrade model we follow implies that the upgrade is 
>>>> possible *Only* from the previous major.minor release. So, the 
>>>> upgrade test for 10.3 will only test any possible upgrades from 10.2.
>>>
>>>
>>>
>>> That makes sense. Thanks for clarifying.
>>
>>
>>
>> Upgrade is possible from any previous release. The origins of the 
>> upgrade test was that any release could be plugged in as the base (to 
>> be upgraded release), even the same version.
>>
>>
>> Dan.
>>
>>
>>
> 

Re: upgrade testing

[Mike Matrigali <mi...@sbcglobal.net>]

I believe that the goal is to be able to upgrade from any release.
I think what you are seeing is that tests haven't caught up with
the upgrade support we want to provide.  I believe we should be
able to hard upgrade a 10.1 db straight to a 10.3 db.

Daniel John Debrunner wrote:
> Bryan Pendleton wrote:
> 
>> Mayuresh Nirhali wrote:
>>
>>> I believe, the upgrade model we follow implies that the upgrade is 
>>> possible *Only* from the previous major.minor release. So, the 
>>> upgrade test for 10.3 will only test any possible upgrades from 10.2.
>>
>>
>> That makes sense. Thanks for clarifying.
> 
> 
> Upgrade is possible from any previous release. The origins of the 
> upgrade test was that any release could be plugged in as the base (to be 
> upgraded release), even the same version.
> 
> 
> Dan.
> 
> 
> 

Re: upgrade testing

[Daniel John Debrunner <dj...@apache.org>]

Bryan Pendleton wrote:
> Mayuresh Nirhali wrote:
>> I believe, the upgrade model we follow implies that the upgrade is 
>> possible *Only* from the previous major.minor release. So, the upgrade 
>> test for 10.3 will only test any possible upgrades from 10.2.
> 
> That makes sense. Thanks for clarifying.

Upgrade is possible from any previous release. The origins of the 
upgrade test was that any release could be plugged in as the base (to be 
upgraded release), even the same version.


Dan.