Cassandra MBeans Reference

This section provides detailed documentation for all major MBeans exposed by Apache Cassandra. Each MBean includes attributes, operations, and practical examples.

MBean Overview

MBean	Domain	Purpose
StorageServiceMBean	o.a.c.db	Core cluster operations
StorageProxyMBean	o.a.c.db	Request coordination
CompactionManagerMBean	o.a.c.db	Compaction control
ColumnFamilyStoreMBean	o.a.c.db	Table operations
GossiperMBean	o.a.c.gms	Gossip protocol
StreamManagerMBean	o.a.c.streaming	Data streaming
CacheServiceMBean	o.a.c.db	Cache management
CommitLogMBean	o.a.c.db	Commit log
HintedHandoffManagerMBean	o.a.c.db	Hint delivery
MessagingServiceMBean	o.a.c.net	Inter-node messaging
EndpointSnitchMBean	o.a.c.locator	Topology
FailureDetectorMBean	o.a.c.gms	Node detection

---

StorageServiceMBean

ObjectName: org.apache.cassandra.db:type=StorageService

The primary MBean for cluster-level operations. Controls node lifecycle, topology changes, and cluster-wide operations.

Key Attributes

Attribute	Type	Description
LiveNodes	List	IP addresses of live nodes
UnreachableNodes	List	IP addresses of unreachable nodes
JoiningNodes	List	Nodes currently joining
LeavingNodes	List	Nodes currently leaving
MovingNodes	List	Nodes currently moving
OperationMode	String	STARTING, NORMAL, JOINING, LEAVING, etc.
ClusterName	String	Cluster name
LocalHostId	String	UUID of this node
Tokens	List	Tokens owned by this node
ReleaseVersion	String	Cassandra version
SchemaVersion	UUID	Current schema version
Keyspaces	List	All keyspace names
NativeTransportRunning	boolean	CQL port status
GossipRunning	boolean	Gossip status
Initialized	boolean	Node initialized

Operations

Operation	Parameters	Description
forceKeyspaceFlush	keyspace	Flush memtables to disk
forceKeyspaceCompaction	keyspace	Trigger major compaction
takeSnapshot	tag, keyspaces	Create snapshot
clearSnapshot	tag, keyspaces	Remove snapshot
repair	keyspace, options	Start repair
decommission	-	Remove node from cluster
move	newToken	Move node to new token
removeNode	hostId	Force remove dead node
assassinate	address	Remove via gossip
drain	-	Flush and stop accepting writes
truncate	keyspace, table	Truncate table
refreshSizeEstimates	-	Update size estimates

nodetool Mappings

nodetool Command	JMX Operation
nodetool status	getLiveNodes(), getUnreachableNodes()
nodetool flush	forceKeyspaceFlush()
nodetool compact	forceKeyspaceCompaction()
nodetool snapshot	takeSnapshot()
nodetool repair	repair()
nodetool decommission	decommission()
nodetool drain	drain()
nodetool info	Various attributes

Example: Get Node Status

ObjectName ssBean = new ObjectName("org.apache.cassandra.db:type=StorageService");

// Get live nodes
List<String> liveNodes = (List<String>) mbsc.getAttribute(ssBean, "LiveNodes");
System.out.println("Live nodes: " + liveNodes);

// Get operation mode
String mode = (String) mbsc.getAttribute(ssBean, "OperationMode");
System.out.println("Mode: " + mode);

// Get cluster name
String cluster = (String) mbsc.getAttribute(ssBean, "ClusterName");
System.out.println("Cluster: " + cluster);

---

StorageProxyMBean

ObjectName: org.apache.cassandra.db:type=StorageProxy

Controls request coordination settings including timeouts and consistency behavior.

Key Attributes

Attribute	Type	Description
TotalHints	long	Total hints stored
HintsInProgress	int	Hints currently being delivered
ReadRepairAttempted	long	Read repair attempts
ReadRepairRepairedBlocking	long	Blocking read repairs
ReadRepairRepairedBackground	long	Background read repairs
SchemaVersions	Map	Schema versions per node

Configuration Attributes

Attribute	Type	Default	Description
RpcTimeout	long	10000ms	General RPC timeout
ReadRpcTimeout	long	5000ms	Read timeout
WriteRpcTimeout	long	2000ms	Write timeout
RangeRpcTimeout	long	10000ms	Range query timeout
CounterWriteRpcTimeout	long	5000ms	Counter write timeout
TruncateRpcTimeout	long	60000ms	Truncate timeout

Operations

Operation	Description
setRpcTimeout	Set general timeout
setReadRpcTimeout	Set read timeout
setWriteRpcTimeout	Set write timeout
reloadTriggerClasses	Reload trigger classes

---

CompactionManagerMBean

ObjectName: org.apache.cassandra.db:type=CompactionManager

Controls compaction operations and provides compaction statistics.

Key Attributes

Attribute	Type	Description
PendingTasks	int	Pending compaction tasks
CompletedTasks	long	Completed compactions
Compactions	List	Currently running compactions
PendingTasksByTableName	Map	Pending tasks per table

Operations

Operation	Parameters	Description
forceUserDefinedCompaction	ks, table, files	Compact specific SSTables
stopCompaction	type	Stop compaction by type
stopCompactionById	compactionId	Stop specific compaction
setConcurrentCompactors	value	Set parallel compactors
getConcurrentCompactors	-	Get parallel compactor count
setCompactionThroughput	mbPerSec	Set throughput limit
getCompactionThroughput	-	Get throughput limit

nodetool Mappings

nodetool Command	JMX Operation
nodetool compactionstats	getCompactions(), getPendingTasks()
nodetool setcompactionthroughput	setCompactionThroughput()
nodetool stop COMPACTION	stopCompaction()

Example: Monitor Compactions

ObjectName cmBean = new ObjectName("org.apache.cassandra.db:type=CompactionManager");

// Get pending compactions
int pending = (int) mbsc.getAttribute(cmBean, "PendingTasks");
System.out.println("Pending compactions: " + pending);

// Get active compactions
List<Map<String, String>> compactions =
    (List<Map<String, String>>) mbsc.getAttribute(cmBean, "Compactions");
for (Map<String, String> c : compactions) {
    System.out.println("Compacting: " + c.get("keyspace") + "." + c.get("columnfamily"));
    System.out.println("  Progress: " + c.get("progress") + "%");
}

---

ColumnFamilyStoreMBean

ObjectName: org.apache.cassandra.db:type=ColumnFamilies,keyspace={ks},columnfamily={table}

Per-table operations and statistics.

Key Attributes

Attribute	Type	Description
LiveSSTableCount	int	Number of live SSTables
LiveDiskSpaceUsed	long	Disk space used (bytes)
TotalDiskSpaceUsed	long	Total disk space
MemtableColumnsCount	long	Columns in memtable
MemtableSwitchCount	int	Memtable flush count
MemtableLiveDataSize	long	Memtable data size
BloomFilterFalsePositives	long	Bloom filter false positives
BloomFilterFalseRatio	double	False positive ratio
MeanPartitionSize	long	Average partition size
MaxPartitionSize	long	Maximum partition size
CompressionRatio	double	Compression ratio
SSTablesPerReadHistogram	long[]	SSTables accessed per read

Operations

Operation	Description
forceFlush	Flush memtable
forceCompactionForTokenRange	Compact token range
estimateKeys	Estimate key count
loadNewSSTables	Load new SSTables from disk
setCompactionStrategyClass	Change compaction strategy
setCompactionThresholds	Set min/max thresholds

---

GossiperMBean

ObjectName: org.apache.cassandra.gms:type=Gossiper

Gossip protocol management.

Key Attributes

Attribute	Type	Description
EndpointStates	Map	State of all known endpoints
DownEndpoints	Set	Down endpoints

Operations

Operation	Description
assassinateEndpoint	Remove endpoint via gossip
unsafeAssassinateEndpoint	Force assassinate
getCurrentGenerationNumber	Get gossip generation
getEndpointDowntime	Get endpoint downtime

---

MessagingServiceMBean

ObjectName: org.apache.cassandra.net:type=MessagingService

Inter-node messaging statistics.

Key Attributes

Attribute	Type	Description
DroppedMessages	Map	Dropped messages by type
RecentlyDroppedMessages	Map	Recently dropped messages
TimeoutsPerHost	Map	Timeouts per remote host

Message Types

Type	Description
MUTATION	Write requests
READ	Read requests
READ_REPAIR	Read repair requests
REQUEST_RESPONSE	Responses
RANGE_SLICE	Range query requests
PAGED_RANGE	Paged range requests
COUNTER_MUTATION	Counter writes
HINT	Hint delivery

Example: Check Dropped Messages

ObjectName msBean = new ObjectName("org.apache.cassandra.net:type=MessagingService");

Map<String, Integer> dropped =
    (Map<String, Integer>) mbsc.getAttribute(msBean, "DroppedMessages");

for (Map.Entry<String, Integer> entry : dropped.entrySet()) {
    if (entry.getValue() > 0) {
        System.out.println("WARNING: " + entry.getValue() +
            " dropped " + entry.getKey() + " messages");
    }
}

---

CacheServiceMBean

ObjectName: org.apache.cassandra.db:type=Caches

Cache management and statistics.

Key Attributes

Attribute	Type	Description
KeyCacheCapacityInBytes	long	Key cache size
KeyCacheSize	long	Keys in cache
KeyCacheHits	long	Cache hits
KeyCacheMisses	long	Cache misses
KeyCacheRecentHitRate	double	Recent hit rate
RowCacheCapacityInBytes	long	Row cache size
RowCacheSize	long	Rows in cache
RowCacheHits	long	Cache hits
RowCacheMisses	long	Cache misses
RowCacheRecentHitRate	double	Recent hit rate

Operations

Operation	Description
invalidateKeyCache	Clear key cache
invalidateRowCache	Clear row cache
setKeyCacheCapacity	Set key cache size
setRowCacheCapacity	Set row cache size
saveCaches	Persist caches to disk

---

Additional MBeans

CommitLogMBean

ObjectName: org.apache.cassandra.db:type=Commitlog

Attribute	Description
CompletedTasks	Completed commit log tasks
PendingTasks	Pending tasks
TotalCommitLogSize	Total commit log size
ActiveSegmentCount	Active segments

HintedHandoffManagerMBean

ObjectName: org.apache.cassandra.db:type=HintedHandoffManager

Attribute	Description
StoredHints	Total stored hints
EndpointsWithPendingHints	Endpoints waiting for hints

StreamManagerMBean

ObjectName: org.apache.cassandra.streaming:type=StreamManager

Attribute	Description
CurrentStreams	Active streams

EndpointSnitchMBean

ObjectName: org.apache.cassandra.locator:type=EndpointSnitchInfo

Attribute	Description
SnitchName	Active snitch class
Datacenter	Local datacenter
Rack	Local rack

FailureDetectorMBean

ObjectName: org.apache.cassandra.gms:type=FailureDetector

Monitors node availability using the Phi Accrual failure detector algorithm.

Attribute	Description
AllEndpointStates	State information for all known endpoints
SimpleStates	Simplified UP/DOWN status for endpoints
PhiValues	Current phi values for each endpoint

Operation	Description
getEndpointState(String)	Get gossip state for specific endpoint
forceConviction(String)	Force mark endpoint as down

---

Quick Navigation

• StorageServiceMBean
• StorageProxyMBean
• CompactionManagerMBean
• ColumnFamilyStoreMBean
• GossiperMBean
• MessagingServiceMBean
• CacheServiceMBean
• StreamManagerMBean
• CommitLogMBean
• HintedHandoffManagerMBean
• EndpointSnitchMBean
• FailureDetectorMBean

---

Next Steps

• Metrics Reference - Detailed metrics documentation
• Connecting to JMX - Setup and security
• Monitoring Guide - Using JMX for monitoring