cassandra - Query-Driven Modelling and Big Data -
i watching 1 of cassandra videos on datasax academy. 1 concept talk lot query driven modelling. makes sense when know queries upfront in killrvideo example.
however, in big data cases, hope not 1 think barely know kind of queries analysts perform on data 5 months or 1 year down road.
if case, best practices storing data? guess advanced querying of such data, end loading data spark. have consider @ storage time avoid operational troubles , troubles @ retrieval time? retrieval approaches less problematic?
cassandra database analytics use cases, not ad-hoc analaytics (only 1 report , query never perform again stuff).
for use cases hadoop cluster better option your. (maybe parquete on hadoop) if see queries perform on , on again, cassandra friend. can use cassandra 50 70% of use cases. column keys , secondary indizies can perform wide spectrum of queries. go analytics guys , ask them need. then: create tables :)
Comments
Post a Comment