Problem Statement
I'm monitoring a Windows Server 2025 host with OpenTelemetry Collector Contrib (v0.150.1) and sending metrics to Dynatrace via OTLP. Most metrics are flowing correctly (CPU, memory, disk, filesystem), but two key issues:
- Health column remains empty (shows -) in the Hosts view
- Process Count and Processes Created metrics are not visible in Dynatrace despite being configured and exported
Configuration
- Platform: Windows Server 2025 Standard
- Collector: otelcontribcol v0.150.1 amd64
- Backend: Dynatrace SaaS ({ENVIRONMENT_ID}.live.dynatrace.com/api/v2/otlp)
- Protocol: OTLP HTTP
Receivers
- hostmetrics/10s — CPU utilization, memory utilization, network, processes count/created, load 1m
- hostmetrics/5m — CPU time, memory usage, disk I/O, filesystem, paging, per-process metrics, load 5m
- hostmetrics/1h — Memory limit, CPU logical/physical counts, uptime, load 15m, processes
- windowsperfcounters/30s — Processor, Memory, Process (7 instances), PhysicalDisk counters
- windows_event_log — System channel
Metrics Being Collected & Exported
✅ system.cpu.utilization (10s) ✅ system.memory.utilization (10s) ✅ system.processes.count (10s, 5m, 1h) ✅ system.processes.created (10s, 5m, 1h) ✅ system.uptime (1h) ✅ system.cpu.load_average.1m/5m/15m (across intervals) ✅ system.memory.limit (1h) ✅ system.cpu.logical.count & system.cpu.physical.count (1h) ✅ All Windows perfcounter metrics (transformed to windows.* namespace) ✅ All system resource attributes (host.name, host.arch, os.type, os.description, os.version, os.build.id, etc.)
Verified: Metrics are reaching Dynatrace (visible in (configured in 3 separate scraper intervals) 2. ✅ Removed cardinality filtering to ensure all metrics pass through 3. ✅ Verified both system.processes.count and system.processes.created metrics are in the config 4. ✅ Verified resource detection is enabled with 13 attributes 5. ✅ Confirmed sending_queue and batch processors are configured 6. ✅ Checked that metrics have correct data types and values 7. ✅ Ensured all metrics flow without errors in collector log with 13 attributes 4. ✅ Confirmed sending_queue and batch processors are configured 5. ✅ Checked that metrics have correct data types and values 6. ✅ Ensured all metrics flow without errors
Current State
- Collector runs without errors
- All metrics export successfully to Dynatrace
- Mey are process count metrics and health status not appearing in Dynatrace?**
For Process metrics: system.processes.count and system.processes.created are configured, exported, and flow without errors. Why don't they appear in Dynatrace UI?
- Missing a required attribute or dimension?
- Wrong metric naming convention?
- Requires a different field/resource mapping?
For Health Status: What does Dynatrace need to calculate and display health status for a Windows host?
- Is there a specific health status metric (system.status, host.health, etc.)?
- Do I need additional metrics beyond what's currently collected?
- Is there a minimum frequency or volume requirement for health calculation?
- Should I be exporting collector internal telemetry (otelcol_* metrics)?
- Is there a specific health status metric (system.status, host.health, etc.)?
- Do I need additional metrics beyond what's currently collected?
- Is there a minimum frequency or volume requirement for health calculation?
- Should I be exporting collector internal telemetry (otelcol_* metrics)?
- Is there a Dynatrace-specific configuration or entity type required?
Configuration Reference
Processors pipeline:
[batch] → [filter (idle CPU)] → [transform (cardinality cleanup)] → [cumulativetodelta] → [metricstransform (perfcounter renaming)] → [resourcedetection] → [otlp_http/dynatrace]
Exporter:
- Endpoint: https://{environmentid}.live.dynatrace.com/api/v2/otlp
- Sending queue: min_size=3000, max_size=3000, flush=60s
Environment
- Dynatrace Version: Latest SaaS ({ENVIRONMENT_ID}.live.dynatrace.com)
- OpenTelemetry Collector: v0.150.1 (Contrib, Windows amd64)
- Collection Running Time: ~30+ minutes (continuous)
Complete Configuration File
otel_dynatrace_only.yaml
extensions: health_check: endpoint: 0.0.0.0:13133 receivers: otlp: protocols: grpc: endpoint: 0.0.0.0:4317 http: endpoint: 0.0.0.0:4318 hostmetrics/10s: collection_interval: 10s scrapers: cpu: metrics: system.cpu.utilization: enabled: true system.cpu.time: enabled: false memory: metrics: system.memory.utilization: enabled: true system.memory.usage: enabled: false network: metrics: system.network.io: enabled: true system.network.packets: enabled: true system.network.errors: enabled: true system.network.dropped: enabled: true system.network.connections: enabled: true processes: metrics: system.processes.count: enabled: true system.processes.created: enabled: true load: metrics: system.cpu.load_average.1m: enabled: true hostmetrics/5m: collection_interval: 5m scrapers: cpu: metrics: system.cpu.time: enabled: true memory: metrics: system.memory.usage: enabled: true disk: metrics: system.disk.io: enabled: true system.disk.operations: enabled: true system.disk.io_time: enabled: true system.disk.operation_time: enabled: true network: metrics: system.network.io: enabled: true system.network.packets: enabled: true system.network.errors: enabled: true system.network.connections: enabled: true system.network.dropped: enabled: true filesystem: include_devices: match_type: strict devices: ["C:", "D:"] metrics: system.filesystem.utilization: enabled: true system.filesystem.inodes.usage: enabled: true paging: metrics: system.paging.usage: enabled: true system.paging.operations: enabled: true process: mute_process_all_errors: true metrics: process.cpu.utilization: enabled: true process.cpu.time: enabled: true process.memory.usage: enabled: true process.memory.virtual: enabled: true process.disk.io: enabled: true processes: metrics: system.processes.count: enabled: true system.processes.created: enabled: true load: metrics: system.cpu.load_average.5m: enabled: true hostmetrics/1h: collection_interval: 1h scrapers: memory: metrics: system.memory.limit: enabled: true cpu: metrics: system.cpu.logical.count: enabled: true system.cpu.physical.count: enabled: true system: metrics: system.uptime: enabled: true load: metrics: system.cpu.load_average.15m: enabled: true processes: metrics: system.processes.count: enabled: true system.processes.created: enabled: true windowsperfcounters: collection_interval: 30s perfcounters: - object: "Processor" instances: ["_Total"] counters: - name: "% Processor Time" - name: "% Privileged Time" - name: "Interrupts/sec" - object: "Memory" counters: - name: "Available MBytes" - name: "% Committed Bytes In Use" - name: "Cache Bytes" - object: "Process" instances: - "svchost" - "lsass" - "csrss" - "services" - "sqlservr" - "w3wp" - "otelcol-contrib" counters: - name: "% Processor Time" - name: "% Privileged Time" - name: "Working Set - Private" - name: "Private Bytes" - name: "Thread Count" - name: "Handle Count" - object: "PhysicalDisk" instances: ["_Total"] counters: - name: "% Disk Time" - name: "Disk Bytes/sec" - name: "Avg. Disk Queue Length" windows_event_log: channel: System start_at: end processors: batch: timeout: 10s send_batch_size: 1024 filter: metrics: datapoint: - metric.name == "system.cpu.utilization" and attributes["state"] == "idle" transform: error_mode: ignore metric_statements: - context: datapoint statements: - delete_key(resource.attributes, "process.cgroup") where IsMatch(metric.name, "^process\\..*") - delete_key(resource.attributes, "process.command") where IsMatch(metric.name, "^process\\..*") - delete_key(resource.attributes, "process.executable.path") where IsMatch(metric.name, "^process\\..*") - delete_key(resource.attributes, "process.owner") where IsMatch(metric.name, "^process\\..*") - delete_key(resource.attributes, "process.parent_pid") where IsMatch(metric.name, "^process\\..*") - delete_key(resource.attributes, "process.command_args") where IsMatch(metric.name, "^process\\..*") - delete_key(datapoint.attributes, "device") where datapoint.attributes["device"] == "" filter/delete-metrics: metric_conditions: - datapoint.attributes["low-memory-process"] != nil cumulativetodelta: max_staleness: 25h metricstransform: transforms: - include: '^\\\\Processor\(_Total\)\\\\% Processor Time$' action: update new_name: windows.processor.time - include: '^\\\\Processor\(_Total\)\\\\% Privileged Time$' action: update new_name: windows.processor.privileged_time - include: '^\\\\Processor\(_Total\)\\\\Interrupts/sec$' action: update new_name: windows.processor.interrupts - include: '^\\\\Memory\\\\Available MBytes$' action: update new_name: windows.memory.available - include: '^\\\\Memory\\\\% Committed Bytes In Use$' action: update new_name: windows.memory.committed_usage - include: '^\\\\Memory\\\\Cache Bytes$' action: update new_name: windows.memory.cache - include: '^\\\\Process\(.*\)\\\\% Processor Time$' action: update new_name: windows.process.cpu_time - include: '^\\\\Process\(.*\)\\\\% Privileged Time$' action: update new_name: windows.process.privileged_time - include: '^\\\\Process\(.*\)\\\\Working Set - Private$' action: update new_name: windows.process.working_set_private - include: '^\\\\Process\(.*\)\\\\Private Bytes$' action: update new_name: windows.process.private_bytes - include: '^\\\\Process\(.*\)\\\\Thread Count$' action: update new_name: windows.process.threads - include: '^\\\\Process\(.*\)\\\\Handle Count$' action: update new_name: windows.process.handles - include: '^\\\\PhysicalDisk\(_Total\)\\\\% Disk Time$' action: update new_name: windows.disk.time - include: '^\\\\PhysicalDisk\(_Total\)\\\\Disk Bytes/sec$' action: update new_name: windows.disk.throughput - include: '^\\\\PhysicalDisk\(_Total\)\\\\Avg\. Disk Queue Length$' action: update new_name: windows.disk.queue_length resourcedetection: detectors: ["system"] system: resource_attributes: host.arch: enabled: true host.id: enabled: true host.name: enabled: true host.ip: enabled: true host.interface: enabled: true host.mac: enabled: true host.cpu.model.name: enabled: true os.type: enabled: true os.description: enabled: true os.name: enabled: true os.version: enabled: true os.build.id: enabled: true exporters: otlp_http/dynatrace: endpoint: "https://{environmentid}.live.dynatrace.com/api/v2/otlp" headers: Authorization: "Api-Token {DYNATRACE_API_TOKEN}" sending_queue: batch: min_size: 3000 max_size: 3000 flush_timeout: 60s service: extensions: [health_check] pipelines: metrics: receivers: [hostmetrics/10s, hostmetrics/5m, hostmetrics/1h, windowsperfcounters] processors: [batch, filter, transform, filter/delete-metrics, cumulativetodelta, metricstransform, resourcedetection] exporters: [otlp_http/dynatrace] logs: receivers: [windows_event_log, otlp] processors: [batch, resourcedetection] exporters: [otlp_http/dynatrace]
Additional Context
The host metrics themselves are healthy and complete. The question is whether Dynatrace requires additional signals, different metric frequencies, or specific configuration to calculate overall host health status.
Any guidance appreciated! 🙏
Featured Posts
