Hello,
Thanks for your answer.
And what do you think about approach of querying data using
OpenTSDB/KairosDB piece by piece, creating a dataframe for each piece,
and then making a union out of them?
This would enable us to store and query data as timeseries and process
it using Spark?
Bes
There are datasource for Cassandra and hbase, however I am not sure how useful
they are, because then you need to do also implement the logic of opentsdb or
kairosdb.
Better to implement your own data sources.
Then, there are several projects enabling timeseries queries in Spark, but I am
not s
