I don't see any issue for you to create pull requests and to start discussions to engage with the community. The main issue, I think, will be to find reviewers able to spend some time testing your contributions. Please make sure there are easy instructions on how to test your code, etc. You'll also need to add a specific Maven profile for these extensions, we cannot afford additional components in the default binary (due to ASF constraints) but the NARs would still be built and published to Maven repositories.
Thanks, Pierre Le mer. 8 sept. 2021 à 17:55, Andres UPM <[email protected]> a écrit : > Hi Pierre, > > Thanks for your response. One of the goals that we can achieve by the use > of NGSI-LD standard is that the users can automate the process of storage. > By using this processor, the users only need to configure some parameters, > regarding the connection to the database and leaving to the processor the > task to create the database, tables, and columns according to the > guidelines provided by the NGSI-LD standard. I understand that make this > for one specific database is not the best option. However, we have a set of > processors for dealing with other databases, SQL and NoSQL but keeping the > same data structure in where the data will be stored. > > With this approach, many users that have implementations using this > standard can use NiFi to store context data in multiple databases without > regarding in define a specific data structure for their application. > > I have known that FIWARE community is very interested in contributing to > this development. Can we start a discussion over this topic?. > > Best Regards, > > Andrés > > > > > El 8 sept 2021, a las 13:29, Pierre Villard <[email protected]> > escribió: > > > > Hi, > > > > I just had a quick look and you propose a processor that is receiving the > > notifications and sending the data into Postgres. Instead, isn't it > better > > to leverage NiFi to receive the notifications, turn these into FlowFiles > > and let users decide what processors they should use next? This way they > > could send the data into many databases, not just Postgres. Or is there a > > reason that makes the data very specific to databases? > > > > Thanks, > > Pierre > > > > Le mar. 7 sept. 2021 à 09:40, Andres UPM <[email protected]> a > écrit : > > > >> Hello there, > >> > >> I am talking on behalf of the New Generation Internet Research gropu of > >> Universidad Politécnica de Madrid and as member of FIWARE Community. In > the > >> last years, we have been working on custom NiFi's processors for > providing > >> NGSI-LD support for data persistence in different storage systems. We > >> already have many processors ready for this purpose and we have the > >> intention to contribute these developments to NiFi project. We have > created > >> the ticket in Jira NIFI-8472 < > >> https://issues.apache.org/jira/browse/NIFI-8472> with more information > >> about it. It is possible that we can start a discussion for analyzing > the > >> possibility of contributing this code? > >> > >> BR, > >> > >> Andrés Muñoz > >
