we have a question about the scalability in particular number of nodes vs amount of ram of each node.
Our problem is the following: we have 8 processes that perform query on elastic and our elastic cluster (o elastic cloud) is composed by two nodes with 8 GB of ram each and we have a query time of 100s for each query. If the processes became 4 or 2 we noticed a decrease of the query time, for 2 precesses query time is about 28 seconds per query.
So our question is: if we would decrement the query time for 8 processes is better to start up another node or increment the amount of ram for the already started nodes?
Thanks a lot for any kind of help.