dhis2-devs team mailing list archive

[Ola Hodne Titlestad <olati@xxxxxxxxxx>]

Using BIRT directly (as Thanh suggests) will of course always be an option,
but then you loose out on the integration of BIRT in DHIS2.
To the users it will be a lot easier when reports are run from within DHIS 2
and part of the DHIS2 dashboard. When using report tables as the data source
for BIRT reports you will loose out on the automatic datamart/report table
generation that happens in the background when you run a BIRT report from
within DHIS2. That means you will have to manually make sure your data
source is updated when using BIRT "standalone".

Since BIRT is very loosely coupled to DHIS2 (you define BIRT's location in
Report Configuration when you start DHIS Reports the first time) I am quite
sure it will be doable to specify a remote central location for BIRT in the
5 "local" DHIS instances using the Report Configuration. This way you will
be able to use the local DHIS instances for both data entry and reports
(which will open in a remote location (URL)).

Lars,
right now we have to specify the absolute path to BIRT on the local machine,
can we alter this to provide the option to specify a URL in stead?


Ola
-------------

On Mon, May 25, 2009 at 4:07 AM, Ngoc Thanh Nguyen <
thanh.hispvietnam@xxxxxxxxx> wrote:

>
>
> On Mon, May 25, 2009 at 3:04 AM, Orvalho Joaquim Augusto <
> orvaquim@xxxxxxxxx> wrote:
>
>> Thanks!
>>
>> This is on discussion here. And for now it sounds too bad.
>>
>> For now it seems that we will have birt on p servers only.
>>
>> Caveman
>
>
> I think it is possible.
> - In BIRT report, you can set the report url refer to any machines (i.e via
> IP address or hostname)
> - You can put report.rpt file in any machines, you access them via http:*
> //sitePrefix*/birt-viewer/ frameset?*option1*=*value1*;*option2*=*
> value2;...
>
> Thanh*
>
>
>
>
>>
>>
>> Jason Pickering wrote:
>>
>>> Not entirely sure I understand your architecture, but I assuming that
>>> all (p) servers will be pointed to the same central database (c). If
>>> you design your BIRT report to point to the (c) server, it will make
>>> no difference whether the report is run from the (p) servers, so long
>>> as the data source points to the (c) server. If you are looking at
>>> doing disconnected data entry, as opposed to a central database, you
>>> will need to collate all of your data into the central server through
>>> import/export. However, the report should still function in the same
>>> way. Just be sure that you design the BIRT report with a data source
>>> to the actual IP address of the server, and not the localhost address.
>>> Obviously, the report will need to be distributed to the (p) servers,
>>> but uploaded without a datasource, as this will need to be specified
>>> in the report itself. Otherwise, perhaps you can just host the report
>>> on the (c) server and have all the (p) servers run it from there by
>>> logging in and running the report remotely.
>>>
>>> Hmm, lots of c's and p's there, but this should certainly be doable.
>>>
>>> Regards,
>>> Jason
>>>
>>>
>>>
>>>
>>> On Sun, May 24, 2009 at 7:55 PM, Orvalho Joaquim Augusto
>>> <orvaquim@xxxxxxxxx> wrote:
>>>
>>>> We plan to have a strange situation.
>>>>
>>>> We will have at least 5 servers (p) for dhis2 data entry. All that data
>>>> will
>>>> be agregated to a central server (c) where will have central reports.
>>>>
>>>> The question: How can I make the dhis on p server to use the birt on c
>>>> server?
>>>>
>>>> It is just an ideia so far. So comments to make it better would be good.
>>>>
>>>> Thanks
>>>> Caveman
>>>>
>>>>