gc-infra/iac/ansible/roles/infra-monitoring/templates/otel-agent-config.yaml.j2

117 lines
2.9 KiB
Django/Jinja

extensions:
# Beobachtet Docker Container
docker_observer:
endpoint: "unix:///var/run/docker.sock"
cache_sync_interval: 30s
receivers:
hostmetrics:
root_path: /hostfs
collection_interval: 15s
scrapers:
cpu:
metrics:
system.cpu.time:
enabled: true
system.cpu.utilization:
enabled: true
memory:
metrics:
system.memory.usage:
enabled: true
system.memory.utilization:
enabled: true
filesystem:
metrics:
system.filesystem.usage:
enabled: true
system.filesystem.utilization:
enabled: true
paging:
metrics:
system.paging.usage:
enabled: true
system.paging.utilization:
enabled: true
system.paging.faults:
enabled: true
load:
disk:
network:
docker_stats:
endpoint: unix:///var/run/docker.sock
collection_interval: 30s
timeout: 20s
# receiver_creator:
# watch_observers: [docker_observer]
# receivers:
# filelog:
# rule: type == "container" # Nur für Container
# config:
# include:
# - /hostfs/var/lib/docker/containers/*/*.log
# operators:
# - type: container
# format: docker
# add_metadata_from_filepath: true
# - type: json_parser
# timestamp:
# parse_from: time
# layout: '%Y-%m-%dT%H:%M:%S.%LZ'
# severity:
# parse_from: stream
# mapping:
# info: stdout
# error: stderr
# Ceph Scraping (Funktioniert nur auf Nodes, wo Ceph Mgr läuft)
prometheus:
config:
scrape_configs:
- job_name: 'ceph-local'
scrape_interval: 30s
scrape_timeout: 10s
static_configs:
- targets: ['127.0.0.1:9283']
metric_relabel_configs:
- source_labels: [__name__]
regex: 'ceph_cluster_total_.*|ceph_health_status|ceph_osd_.*|ceph_pool_.*'
action: keep
processors:
batch:
timeout: 5s
resourcedetection:
detectors: [env, system]
resourcedetection/docker:
detectors: [env, docker]
timeout: 2s
override: false
exporters:
debug:
verbosity: detailed
otlp:
endpoint: "127.0.0.1:4317"
headers:
authorization: {{ hyperdx_api_ingestion_key }}
compression: gzip
tls:
insecure: true
service:
extensions: [docker_observer]
pipelines:
metrics:
receivers: [hostmetrics, docker_stats, prometheus]
# receivers: [hostmetrics]
processors: [resourcedetection, batch]
exporters: [otlp]
# logs:
# receivers: [receiver_creator]
# processors: [resourcedetection/docker, batch]
# exporters: [otlp, debug]