Thursday, December 26, 2013

cassandra 2.0 catch 101 – part5

Our cluster status.
Datacenter: datacenter1
|/ State=Normal/Leaving/Joining/Moving
-- Address Load Tokens Owns Host ID Rack
UN 147.34 KB 512 34.3% f13c9390-4c52-4fc2-afa8-f7f74e7fd710 rack1
UN 123.31 KB 512 33.2% bc7fcfcc-9a30-4929-bf24-35ec770856a3 rack1
UN 160.74 KB 256 16.5% 999d58bf-2b31-49ff-a452-6f0d01598429 rack1
UN 137.39 KB 256 16.0% 222796e9-d330-469a-8dcd-3f3581c9d795 rack1

So it is pretty interesting that a node can own different amount of cluster load based on the tokens specified. Because in our cluster environment, we have different types of hardware and for instance, mine is pretty old. With default settings, a default of 512 tokens is assigned.

The setting for tokens can be found in cassandra.yaml
# This defines the number of tokens randomly assigned to this node on the ring
# The more tokens, relative to other nodes, the larger the proportion of data
# that this node will store. You probably want all nodes to have the same number
# of tokens assuming they have equal hardware capability.
# If you leave this unspecified, Cassandra will use the default of 1 token for legacy compatibility,
# and will use the initial_token as described below.
# Specifying initial_token will override this setting.
# If you already have a cluster with 1 token per node, and wish to migrate to
# multiple tokens per node, see
num_tokens: 512

Just note that this setting is one time only when a join the cluster. Meaning that the first time it join, 512 tokens will be assign for this node and this tokens will be store in the keyspace system and in column family local. Even if you removed the data directory and start all over, the data will be stream from other nodes, hence, the information is still persists. If you really want to change at later day, it is possible, you may want to treat this node as dead through decommission, stop cassandra instance. Change the num_tokens configuration in yaml file and then start cassandra instance back. You may want to think about this because decommission stream data to other servers and it may create load in system and also network traffic.

No comments:

Post a Comment