Hi Dominik,

Ok. We also need a strategy on deleting data from the data lake.
I think it makes sense to define a retention time per persisted stream.

Additionally we should discuss when the data is deleted. 
Currently we only have a beta feature to clean all data in the database, which 
is not sufficient.

One of the main questions is: Should the data be kept in the database when the 
pipeline / source is deleted?
From a user point of view it might make sense to remove all corresponding data 
when deleting a pipeline. But I often use this feature to collect some (high 
frequency) data for some time and remove the pipeline, but I want to keep the 
data to analyze it later (e.g. train a machine learning model).

Any thoughts on that?

Philipp

On 2021/01/18 21:56:21, Dominik Riemer <[email protected]> wrote: 
> Hi Philipp,
> 
> to be honest, I don't know yet - but I'd say there will be some data model 
> that links a data stream to the corresponding index. This is probably similar 
> to how we currently handle "visualizable pipelines".
> 
> Dominik
> 
> On 2021/01/18 21:35:20, Philipp Zehnder <[email protected]> wrote: 
> > Hi,
> > very cool, the API is now much more clear.
> > 
> > How will we handle the read of data from the data lake in the future?
> > Currently the index name is used, but I guess we will change this to the 
> > stream id,
> > right? How will this look like?
> > 
> > Philipp
> > 
> > Am 18.01.2021 um 08:33 schrieb Dominik Riemer <[email protected]>:
> > 
> > Hi all,
> > 
> > 
> > 
> > I did some changes to the data explorer management and extracted all queries
> > for historic data to an individual module (streampipes-data-explorer). While
> > doing this, I also split the previously existing DataLakeManagementv3 class
> > into several smaller units to make it more readable.
> > 
> > The change is currently in the STREAMPIPES-272 branch [1], but I think we
> > can merge this soon for the next release.
> > 
> > 
> > 
> > Dominik
> > 
> > 
> > 
> > [1]
> > https://github.com/apache/incubator-streampipes/tree/STREAMPIPES-272/streamp
> > ipes-data-explorer/src/main/java/org/apache/streampipes/dataexplorer
> > 
> > 
> > 
> > 
> 

Reply via email to