CODEWITHSUNDEEP
parallel-processingcassandracluster-computing
user2924472
user2924472

Reputation: 25

Cassandra - Optimizing hardware in cluster

I have been able to get Cassandra working on a macbook cluster (for fun). Now I am trying to operationalize this for research.

Currently, I have a single linux machine running intel 3770K lga 1150. I would like to create a cluster for the purpose on running cassandra. Can I use cheap machines (2-3 nodes with intel i5, 4tb hd, and 8 gb ram)? What is the best configuration to do this right the first time?

Is is possible to use the new nodes to operate cassandra and the current machine just utilize the data for analysis?

Upvotes: 0

Views: 202

Answers (1)

Chris Lohfink
Chris Lohfink

Reputation: 16400

8gb ram is pretty low. Id recommend a minimum of 16gb (more the better) so you can safely allocate 8gb heap while leaving room for the offheap stuff. Especially if you want to store multiple TB of data on it you want more then 8gb. Some data models are worse then others. If using non-ssd's be sure to have a dedicated drive to the commitlog so its not competing with data. It will work with what you listed but you wont get good performance once theres a decent amount of data.

http://www.datastax.com/documentation/cassandra/2.0/cassandra/architecture/architecturePlanningAbout_c.html

You can create multiple data centers to separate your different workloads. DSE workload snitch will do that for you if using datastax enterprise.

Upvotes: 0

Related Questions