For example if I process a lot of transactions [timestamp (key), value, trader,..] I would like to have random partitioning for the sake of load balancing.
But for the sake of queries I would like these transactions to be ordered by timestamps on each node so that it would be very fast to go through all transactions in the last 5 minutes (or some other period) and combine totals using MapReduce.
Why does cassandra sort keys in the file randomly if I use random partitioner?
Q1: Is there a workaround for this? I could use indexing but it is not ideal.
Q2. Out of curiosity I would like to know why is key sorting in one node related to partition strategy?