Skip to main content

Exposed metrics

The following tables describe the metrics exported by OBI in both OpenTelemetry (OTEL) and Prometheus formats, for use within Coralogix observability workflows.

Application metrics

Changed in v0.11.0
FamilyName (OTEL)Name (Prometheus)TypeUnitDescription
Applicationhttp.client.request.durationhttp_client_request_duration_secondsHistogramsecondsTime taken for outbound HTTP calls from client applications
Applicationhttp.client.request.body.sizehttp_client_request_body_size_bytesHistogrambytesSize of HTTP request payloads sent by clients
Applicationhttp.client.response.body.sizehttp_client_response_body_size_bytesHistogrambytesSize of HTTP responses received by clients
Applicationhttp.server.request.durationhttp_server_request_duration_secondsHistogramsecondsTime taken to process inbound HTTP requests on servers
Applicationhttp.server.request.body.sizehttp_server_request_body_size_bytesHistogrambytesSize of incoming HTTP request bodies received by servers
Applicationhttp.server.response.body.sizehttp_server_response_body_size_bytesHistogrambytesSize of outbound HTTP response bodies sent by servers
Applicationrpc.client.durationrpc_client_duration_secondsHistogramsecondsDuration of outbound gRPC service calls
Applicationrpc.server.durationrpc_server_duration_secondsHistogramsecondsDuration of inbound RPC service calls
Applicationsql.client.durationsql_client_duration_secondsHistogramsecondsTime taken for SQL client operations (Experimental)
Applicationredis.client.durationredis_client_duration_secondsHistogramsecondsTime taken for Redis client operations (Experimental)
Applicationmessaging.publish.durationmessaging_publish_durationHistogramsecondsTime taken for message publication (e.g., Kafka) (Experimental)
Applicationmessaging.process.durationmessaging_process_durationHistogramsecondsTime taken for message processing (e.g., Kafka) (Experimental)
Application Processprocess.cpu.timeprocess_cpu_time_seconds_totalCountersecondsTotal CPU time consumed, broken down by state (system/user/wait)
Application Processprocess.cpu.utilizationprocess_cpu_utilization_ratioGaugeratioCPU time usage change over time, normalized by CPU count
Application Processprocess.memory.usageprocess_memory_usage_bytesUpDownCounterbytesPhysical memory currently used
Application Processprocess.memory.virtualprocess_memory_virtual_bytesUpDownCounterbytesVirtual memory committed
Application Processprocess.disk.ioprocess_disk_io_bytes_totalCounterbytesTotal disk I/O bytes transferred
Application Processprocess.network.ioprocess_network_io_bytes_totalCounterbytesTotal network I/O bytes transferred
Go runtimego.memory.limitgo_memory_limit_bytesUpDownCounterbytesGo runtime memory limit. Added in OBI v0.10.0.
Go runtimego.memory.gc.goalgo_memory_gc_goal_bytesUpDownCounterbytesCommitted GC heap goal. Added in OBI v0.11.0. Omitted when the runtime offset and the fallback symbol are both unavailable.
Go runtimego.memory.gc.cyclesgo_memory_gc_cycles_totalCountercountCompleted Go GC cycles. Added in OBI v0.10.0.
Go runtimego.memory.gc.pause.durationgo_memory_gc_pause_duration_secondsHistogramsecondsCumulative stop-the-world GC pause duration. Added in OBI v0.11.0. Requires Go 1.22+.
Go runtimego.memory.usedgo_memory_used_bytesUpDownCounterbytesMemory in use, broken down by go.memory.type (stack or other). Added in OBI v0.11.0. Requires Go 1.23+.
Go runtimego.memory.allocatedgo_memory_allocated_bytes_totalCounterbytesCumulative allocated heap bytes. Added in OBI v0.11.0. Requires Go 1.23+.
Go runtimego.memory.allocationsgo_memory_allocations_totalCounterallocationsCumulative heap allocation count. Added in OBI v0.11.0. Requires Go 1.23+.
Go runtimego.cpu.timego_cpu_time_seconds_totalCountersecondsCumulative CPU time, broken down by go.cpu.state (and go.cpu.detailed_state where applicable). Added in OBI v0.11.0. Requires Go 1.23+.
Go runtimego.goroutine.countgo_goroutine_countUpDownCountergoroutinesCurrent goroutine count. Added in OBI v0.11.0. Omitted rather than reported partially when the runtime snapshot is incomplete.
Go runtimego.processor.limitgo_processor_limitUpDownCounterprocessorsGOMAXPROCS value. Added in OBI v0.10.0.
Go runtimego.config.gogcgo_config_gogc_percentUpDownCounterpercentGOGC configuration value. Added in OBI v0.10.0.
Go runtimego.schedule.durationgo_schedule_duration_secondsHistogramsecondsCumulative runnable-to-running goroutine latency. Added in OBI v0.11.0. Requires Go 1.20+.
JVMjvm.memory.usedjvm_memory_used_bytesUpDownCounterbytesHotSpot JVM memory used, per memory pool. Added in OBI v0.10.0. Sum by jvm.memory.type for heap and non-heap totals.
JVMjvm.memory.committedjvm_memory_committed_bytesUpDownCounterbytesHotSpot JVM memory committed. Added in OBI v0.10.0.
JVMjvm.memory.limitjvm_memory_limit_bytesUpDownCounterbytesHotSpot JVM memory limit. Added in OBI v0.10.0.
JVMjvm.memory.used_after_last_gcjvm_memory_used_after_last_gc_bytesUpDownCounterbytesHotSpot JVM memory used after the last GC. Added in OBI v0.10.0.
Networkobi.network.flow.bytesobi_network_flow_bytesCounterbytesBytes transmitted between source and destination network endpoints
Networkobi.network.flow.packetsobi_network_flow_packets_totalCountercountPackets transmitted between source and destination network endpoints. Added in OBI v0.10.0.
Networkobi.network.inter.zone.bytesobi_network_inter_zone_bytesCounterbytesCross-zone network bytes within Kubernetes clusters (Experimental)
Networkobi.stat.tcp.ioobi_stat_tcp_io_bytes_totalCounterbytesTCP bytes transferred. Added in OBI v0.10.0.
Networkobi.stat.tcp.retransmitsobi_stat_tcp_retransmits_totalCountercountTCP segment retransmissions. Added in OBI v0.10.0. The Prometheus name gained the _total suffix in OBI v0.11.0.
Networkobi.stat.tcp.failed.connectionsobi_stat_tcp_failed_connections_totalCounterconnectionsTCP failed connection count, labeled by network.tcp.handshake.role (client or server) and failure reason. Added in OBI v0.9.0. The reason label is disabled by default. The Prometheus name gained the _total suffix in OBI v0.11.0.

Go runtime and JVM metrics are both enabled through the shared application_runtime metrics feature. From OBI v0.11.0 that is the only way to enable them: the application_jvm alias and the jvm_runtime_metrics.enabled flag were removed, though jvm_runtime_metrics.sampling_interval remains as a sampling control.

OBI also supports exporting Span metrics and Service Graph metrics, which can be toggled via feature configuration options.

From OBI v0.11.0, Prometheus metric names are derived from their OTLP definitions rather than declared separately, which corrected several of them — see the _total suffix notes in the table above and the CUDA GPU renames in What's changed. On the OTLP export path, target_info, traces_target_info, traces_host_info, traces_span_metrics_calls_total, and traces_span_metrics_duration are now emitted in dot notation as target.info, traces.target.info, traces.host.info, traces.span.metrics.calls, and traces.span.metrics.duration; the Prometheus scrape path keeps the underscore names. The OBI-specific obi.jvm.heap.used metric (obi_jvm_heap_used_bytes) was removed — sum jvm.memory.used by jvm.memory.type for heap totals instead.

From OBI v0.10.0, HTTP spans emit url.query by default when present, and client HTTP spans include the query string in url.full. Sensitive parameters are automatically redacted; extend the redaction list with OTEL_EBPF_SENSITIVE_QUERY_PARAMS_ADD. Java route harvesting is disabled by default, so the http.route attribute on Java spans falls back to the raw URL path. See What's changed in the overview.

From OBI v0.9.0, database error response bodies are no longer included on SQL spans by default. If you previously relied on sql.client.duration spans carrying database error responses, enable that capture explicitly. See What's changed in the overview.

From OBI v0.7.0+, DNS calls are excluded from span metrics and service-graph metrics by default. SQL operations (MySQL, PostgreSQL, MariaDB) are now reported as SERVER spans, so any aggregations or filters that key on span kind for SQL traffic should be updated accordingly. See What's changed in the overview for details.

Attribute selection for OBI metrics

Metrics and attributes follow OTEL dot.notation but are converted to Prometheus underscore_notation in Prometheus outputs.

You can customize which attributes are exposed or hidden by using the attributes->select section in the configuration.

Common attributes table

MetricsAttribute NameDefault
Application (all)http.request.methodshown
Application (all)http.response.status_codeshown
Application (all)http.routeshown if routes config exists
Application (all)k8s.daemonset.nameshown if Kubernetes metadata is on
Application (all)k8s.deployment.nameshown if Kubernetes metadata is on
Application (all)k8s.namespace.nameshown if Kubernetes metadata is on
Application (all)k8s.node.nameshown if Kubernetes metadata is on
Application (all)k8s.owner.nameshown if Kubernetes metadata is on
Application (all)k8s.pod.nameshown if Kubernetes metadata is on
Application (all)k8s.container.nameshown if Kubernetes metadata is on
Application (all)k8s.pod.start_timeshown if Kubernetes metadata is on
Application (all)k8s.pod.uidshown if Kubernetes metadata is on
Application (all)k8s.replicaset.nameshown if Kubernetes metadata is on
Application (all)k8s.statefulset.nameshown if Kubernetes metadata is on
Application (all)k8s.cluster.nameshown if Kubernetes metadata is on
Application (all)service.nameshown
Application (all)service.namespaceshown
Application (all)target.instanceshown
Application (all)url.pathhidden
Application (client)server.addresshidden
Application (client)server.porthidden
Application (process)various process-specific attributes (e.g., process.command, process.executable.path)shown if process metrics enabled
Application RPCvarious RPC attributes (rpc.method, rpc.grpc.status_code)shown
Application (server)client.addresshidden
Networknetwork-specific attributes (e.g., src.address, dst.address, iface)mostly hidden or conditional
Network (TCP failed connections)network.tcp.handshake.role — values client or servershown

Note: The obi.network.inter.zone.bytes metric shares the same attribute set as obi.network.flow.bytes but all are hidden by default except k8s.cluster.name, src.zone, and dst.zone.

Internal OBI metrics

Changed in v0.11.0

OBI provides several internal Prometheus-format metrics to help monitor its own health and activity:

NameTypeDescription
obi_ebpf_tracer_flushesHistogramSize of trace groups flushed from eBPF to pipeline
obi_otel_metric_exports_totalCounterTotal metric batches sent to remote OTEL collector
obi_otel_metric_export_errors_totalCounterVecCount of OTEL metric export errors by type
obi_otel_trace_exports_totalCounterTotal trace batches sent to remote OTEL collector
obi_otel_trace_export_errors_totalCounterVecCount of OTEL trace export errors by type
obi_prometheus_http_requests_totalCounterVecPrometheus scrape endpoint request count by port and path
obi_instrumented_processesGaugeVecNumber of processes instrumented by Obi, labeled by process name
obi_instrumentation_errors_totalCounterVecInstrumentation errors, labeled by process executable name and error type
obi_avoided_servicesGaugeVecServices OBI skipped because they already carry OpenTelemetry instrumentation, labeled by service name, namespace, and telemetry type. Emitted only when the avoided-services internal metric is enabled.
obi_internal_build_infoGaugeVecObi version information including build timestamp and commit ID

From OBI v0.11.0, obi_otel_trace_exports_total and obi_otel_trace_export_errors_total reflect the real outcome reported by the exporter instead of counting attempts, so error counts that previously read as zero may become non-zero without anything else changing. Internal metrics also carry host metadata as resource attributes.

OBI can additionally report eBPF probe latency and map utilization, but that family is opt-in — it requires the ebpf metrics feature with a Prometheus endpoint, or internal_metrics.bpf_metric_scrape_interval set above zero. See the upstream BPF metrics collection reference for the metric names and labels.

Last updated on
Includes latestOlder versionDeprecated
On this page
Was this page helpful?