cassandra_node_quiesce

#!/usr/bin/env bash

# Use nodetool to disable Cassandra only on a specific node.

# There are times when you would like a Cassandra node to stop receiving requests or appear like it is down to outside clients.
# You may want to use this to disable a node and adjust your log4j properties for more debugging,
# or your may be running some administrative action like repair or rebuildsstables and you want this action to complete as quickly as possible. 
# Being able to disable Cassandra without shutting it down completely can be very handy.

# How it works:
# Gossip is the protocol that Cassandra nodes use to determine which nodes are UP or DOWN.
# Cassandra nodes gossip to each other and maintain a state of the cluster.
# By disabling gossip other Cassandra nodes will stop forwarding read and write requests to this node 
# because they will believe the node is down. Client using thrift will still be able to connect to a node in this state.
# Disablebinary disables the client port 9042 which prevents clients from connecting to this node. 
# Disablethrift disables the client port 9160 which prevents clients from connecting to this node. 

trace_cmd=true
cassandra_nodetool disablegossip "$@"		# Disable gossip (effectively marking the node down)
cassandra_nodetool disablethrift "$@"		# Disable thrift server
cassandra_nodetool disablebinary "$@"		# Disable native transport (binary protocol)

# cassandra_nodetool disableautocompaction "$@"	# Disable autocompaction for the given keyspace and table
# cassandra_nodetool disablebackup "$@"		# Disable incremental backup
# cassandra_nodetool disablehandoff "$@"	# Disable storing hinted handoffs
# cassandra_nodetool disablehintsfordc "$@"	# Disable hints for a data center
# warning: test and add disabling streaming and binary protocol if needed

cassandra_nodetool drain "$@"			# In theory, this should be the only request needed, but different versions have different issues justifying the rest of the lines in this file.
# sleep 300					# If flush is interrupted, manually perform sleep for 5 minutes to wait for all data to get flushed.

cassandra_nodetool flush "$@"			# Empty the memtables and pending commit log changes into sstables
# sleep 300					# If flush is interrupted, manually perform sleep for 5 minutes to wait for all data to get flushed.

sync; sync					# Force Linux all cached file data to disk. The second sync waits.