Helm Values

This page summarizes user-facing values. The chart values.yaml files remain the executable defaults.

Collector

Path	Default	Purpose
image.repository	otel/opentelemetry-collector-contrib	Collector image
image.tag	0.156.0	Collector tag
image.digest	empty	Optional immutable digest
router.replicaCount	2	Fixed stateless router count
router.queueSize	100000	Router-to-backend sending queue
backend.mode	singleWriter	singleWriter or explicit horizontal capacity mode
backend.replicaCount	1	Must be one in singleWriter; at least two in horizontal
backend.serviceGraph.storeTtl	10s	Unpaired span retention
backend.serviceGraph.storeMaxItems	10000	Pairing-store limit
backend.serviceGraph.metricsFlushInterval	30s	Metric flush period
backend.serviceGraph.metricBatchSize	256	Maximum datapoints per Kafka-bound batch
streamContract.kafka.brokers	example broker	Kafka bootstrap servers
streamContract.kafka.security.protocol	SASL_SSL	PLAINTEXT, SASL_PLAINTEXT, or SASL_SSL
streamContract.kafka.security.existingSecret	servicegraph-kafka-auth	Existing SASL credentials
streamContract.topics.servicegraphMetrics	otel.servicegraph.metrics	Metrics topic
rootSpanDiscovery.enabled	false	Add root filtering and the spanmetrics discovery pool
rootSpanDiscovery.backend.replicaCount	2	Identity-sharded spanmetrics backends
rootSpanDiscovery.backend.metricsFlushInterval	60s	Discovery delta flush period
rootSpanDiscovery.backend.metricsExpiration	2m	Inactive metric resource expiration
rootSpanDiscovery.backend.seriesExpiration	2m	Inactive series expiration
rootSpanDiscovery.backend.aggregationCardinalityLimit	0	Unlimited identities; avoids overflow collapse
rootSpanDiscovery.backend.metricBatchSize	256	Maximum discovery datapoints per Kafka-bound batch

Resource, scheduling, ServiceAccount, image-pull, and security-context values are also available in the chart.

Flink

Path	Default	Purpose
image.ref	internal example	Complete runtime image
application.clusterId	servicegraph-diff	Standalone Flink cluster ID
application.parallelism	3	Job parallelism
application.jobManagerReplicas	1	Restartable JobManager count
application.taskManagerReplicas	2	Fixed TaskManager count
application.taskManagerSlots	2	Slots per TaskManager
job.fixedJobId	000...001	Stable Flink job ID
job.allowNonRestoredState	false	Allow reviewed upgrades to discard unmapped savepoint state
job.groupId	graph-element-engine	Kafka source group
job.interactionTtlSeconds	86400	Delete inactivity threshold
job.elementTtlSeconds	{}	Optional inactivity thresholds by semantic element type
job.contributionCoalescingIntervalSeconds	300	Minimum interval for unchanged contributor refreshes; 0 disables suppression
job.allowedLatenessSeconds	60	Watermark out-of-order bound
job.checkpointIntervalMs	30000	Checkpoint interval
job.checkpointTimeoutMs	600000	Checkpoint timeout in milliseconds
job.checkpointMinPauseMs	5000	Pause between completed checkpoints
job.tolerableFailedCheckpoints	1	Consecutive asynchronous failures tolerated
job.restartAttempts	3	Fixed-delay attempts
state.backend	rocksdb	rocksdb or explicit hashmap fallback
state.incrementalCheckpoints	true	Incremental RocksDB checkpoints
state.rocksdb.localDirectory	/flink-rocksdb	TaskManager-local RocksDB working directory
state.rocksdb.localStorageSize	10Gi	RocksDB emptyDir size limit
resources.taskManager.requests.ephemeral-storage	1Gi	TaskManager ephemeral-storage request
resources.taskManager.limits.ephemeral-storage	12Gi	TaskManager ephemeral-storage limit
logging.rootLevel	INFO	Flink console log level
rbac.create	true	Create ConfigMap-only runtime RBAC
storage.createClaim	true	Create state claim
storage.existingClaim	empty	Use existing state claim
storage.storageClassName	rwx	Shared storage class
storage.accessModes	[ReadWriteMany]	State volume access
storage.retainClaim	true	Keep created claim on uninstall

Kafka security fields and every enabled topic name must match the Collector and consumers.

ArangoDB indexer

Path	Default	Purpose
image.repository	internal example	Indexer image
image.tag	0.1.0	Indexer image tag
replicas	1	Kafka indexer replicas
consumerGroupId	servicegraph-arangodb-indexer	Kafka group
arangodb.urls	local example	ArangoDB endpoints
arangodb.database	servicegraph	Existing database
arangodb.graph	servicegraph	Named graph
arangodb.allowDatabaseCreation	false	Local-only database creation
arangodb.auth.existingSecret	writer Secret	Writer credentials

Gremlin Server

Path	Default	Purpose
image.repository	internal example	Validated Gremlin image
image.tag	0.1.0	Gremlin image tag
replicas	1	Read-only server replicas
service.port	8182	GraphBinary/WebSocket port
gremlin.evaluationTimeoutMs	30000	Traversal timeout
arangodb.auth.existingSecret	reader Secret	Read-only credentials
jvm.options	bounded 384 MiB heap	JVM memory settings

Local ArangoDB

Path	Default	Purpose
image.tag	3.12.9.4	Local ArangoDB version
persistence.size	2Gi	Development data claim
persistence.existingClaim	empty	Optional existing RWO claim

Demo

Path	Default	Purpose
collector.endpoint	router OTLP HTTP URL	Trace destination
traffic.emitIntervalSeconds	2	Delay between batches
traffic.topologyChangeIntervalSeconds	20	Delay between graph changes
traffic.initialEdges	2	Initial active edges
traffic.maxActiveEdges	6	Maximum active edges
traffic.requestsPerTick	3	Traces per batch
traffic.errorRate	0.08	Failed trace probability
traffic.serviceNamespace	shop	Emitted service namespace
traffic.instanceId	live-demo	Instance-ID suffix
traffic.randomSeed	empty	Optional deterministic seed

Inspect exact defaults

helm show values deploy/helm/servicegraph-collector
helm show values deploy/helm/servicegraph-flink
helm show values deploy/helm/servicegraph-arangodb
helm show values deploy/helm/servicegraph-indexer
helm show values deploy/helm/servicegraph-gremlin
helm show values deploy/helm/servicegraph-demo