Bringing together the Apache Cassandra experts from the community and DataStax.
Want to learn? Have a question? Want to share your expertise? You are in the right place!
Not sure where to begin? Getting Started
Hi!
Looking into Cassandra configuration options I see `nodetool setstreamthroughput` which is explained as "Sets the throughput capacity in megabits per second (Mb/s) for outbound streaming in the system".
I'm not sure what means "outbound streaming in the system". Does it affect the performance of query result output from Cassandra node? When this setting needs to be changed?
Operations such as bootstrapping and repairs stream parts of SSTables between nodes. Setting the streaming throughput on nodes determines how fast data is streamed by the node to other nodes in the cluster.
A high streaming throughput will have significantly higher IO and will affect other operations on the node including read requests because these operations ultimately are competing for the same IO bandwidth of the disk and network.
If you are adding a node and want it to bootstrap quicker, you would increase the streaming throughput of the other nodes in the cluster but understand that there is a finite amount of resources (CPU, RAM, disk IO, network bandwidth) on each node and you are allocating more resources to one operation (streaming) at the expense of another (read requests).
Only you can determine how much throughput to allocate based on trial and error. Cheers!
6 People are following this question.
