jsr338-experts@jpa-spec.java.net

[jsr338-experts] Re: support for multitenancy

From: Linda DeMichiel <linda.demichiel_at_oracle.com>
Date: Tue, 10 Apr 2012 14:27:02 -0700

On 4/10/2012 2:25 PM, michael keith wrote:
> Er.. just to clarify...
> We started out talking about a per-thread tenant id for SaaS (to be offered in the future) and am I right that somehow
> the impression was given that that tenant id will be in JNDI or passed in EE 7? As it stands right now (in EE 7),
> containers are not going to pass in (or post in JNDI) the tenant id of a per-thread (SaaS) tenant. They will only be
> expected to maintain a single tenant id for the life of an application instance, so this won't help for SaaS.
>

Right. This is not expected to be on a per-thread basis in Java EE 7.


> On 10/04/2012 1:19 PM, Linda DeMichiel wrote:
>>
>>
>> On 4/10/2012 10:16 AM, Steve Ebersole wrote:
>>> On 04/10/2012 11:33 AM, Linda DeMichiel wrote:
>>>>
>>>>
>>>> On 4/10/2012 9:26 AM, Steve Ebersole wrote:
>>>>>>>>> Ideally, the management of the tenant id should be transparent to the
>>>>>>>>> application, although we should revisit this in Java EE 8 as we move
>>>>>>>>> further into support for SaaS.
>>>>>>>>
>>>>>>>> For the application to not have to manage tenant ids, I guess the
>>>>>>>> tenant identifier
>>>>>>>> would need to be available to the provider on a per-invocation basis
>>>>>>>> (in a thread
>>>>>>>> context set by the container)? As you mention, not something that we
>>>>>>>> necessarily have to worry about now, but just so we know what we will
>>>>>>>> need
>>>>>>>> in the future if this is what we want.
>>>>>>>>
>>>>>>>
>>>>>>> Right. For shared application instances we will need this.
>>>>>>
>>>>>> You had said this would be available via ENC/JNDI before I think?
>>>>>
>>>>> Will the tenant identifier be available in ENC/JNDI regardless of
>>>>> strategy being used?
>>>>>
>>>>
>>>> Yes
>>>
>>> And will it be available from JNDI when the EMF is being built?
>>
>> I wasn't assuming that. I was assuming that the container would pass in the
>> requisite information to the createEMF call, as we do today.
>>