Set up Traces Collection for Kubernetes Environments

After installing or upgrading your Sumo Logic Kubernetes Collection, you will be able to send your traces directly to its endpoint using OpenTelemetry (as well as older formats like Jaeger or Zipkin).

Traces will be enhanced with Kubernetes metadata, similarly to the logs and metrics collected by the collector. See below for installation instructions.

Prerequisites

• Kubernetes 1.24+
• Helm 3.5+

Account Type	Account Level
Credits	Enterprise Operations and Enterprise Suite Essentials get up to 5 GB a day

Installing Sumo Logic Tracing on Kubernetes

Installation is the same as for the official Sumo Logic Kubernetes Collection. The process uses a Helm chart to set all required components. It will automatically download and configure the OpenTelemetry Collector, which will collect, process, and export telemetry data to Sumo Logic.

In the following installation steps, we use the release name collection and the namespace name sumologic. You can use any names you want, however, you'll need to adjust your installation commands to use your names since these names impact the OpenTelemetry Collector endpoint name.

note

If you're upgrading from Sumo Logic Kubernetes Collection:

• v2.x to v3.x, see Sumo Logic Kubernetes Collection Migration Guide v3
• v3.x to v4.x, see Sumo Logic Kubernetes Collection Migration Guide v4

Collection architecture

Tracing data from your services is sent through multiple local OpenTelemetry Collectors/Agents, deployed as a StatefulSet (otelcol-instrumentation), which buffers and sends data to a OpenTelemetry Collector gateway (traces-gateway). Finally, the data is sent to the OpenTelemetry Collector (traces-sampler) helping to shape and trim the traffic, both running as a Deployment.

Installing or upgrading to the latest version

Refer to install/upgrade instructions for the current version. Tracing is enabled by default.

If you plan to auto-instrument your Java, Python, and JS applications in K8s environments, use the Helm command in that article.

Using the command line

helm upgrade --install collection sumologic/sumologic \
  --namespace sumologic \
  --create-namespace \
  --set sumologic.accessId=<SUMO_ACCESS_ID> \
  --set sumologic.accessKey=<SUMO_ACCESS_KEY> \
  --set sumologic.clusterName="<MY_CLUSTER_NAME>"

Using a configuration file

We recommend creating a new values.yaml file for each Kubernetes cluster you wish to install collection on and setting only the properties you wish to override. For example:

sumologic:
  accessId: <SUMO_ACCESS_ID>
  accessKey: <SUMO_ACCESS_KEY>
  clusterName: <MY_CLUSTER_NAME>

Once you've customized your config, you can use the following commands to install or upgrade:

helm upgrade --install collection sumologic/sumologic \
  --namespace sumologic \
  --create-namespace \
  -f values.yaml

Enabling tracing for existing installations

If you previously installed sumologic-kubernetes-collection v2.x without enabling tracing, it can be enabled with sumologic.traces.enabled=true.

Using command line

helm upgrade --install collection sumologic/sumologic \
  --namespace sumologic \
  ... \
  --set sumologic.traces.enabled=true

Using a configuration file

The values.yaml file needs to have the relevant section enabled, such as:

sumologic:
  ...
  traces:
    enabled: true

After updating the configuration file, the changes can be applied with the following:

helm upgrade --install collection sumologic/sumologic \
  --namespace sumologic \
  -f values.yaml

Custom configuration

Every customization to your collector configuration should be applied using keys config.merge or config.override:

• For otelcolInstrumentation. Use otelcolInstrumentation.config.merge, which will merge provided configuration to default configuration or otelcolInstrumentation.config.override, which will override the whole configuration.
• For tracesGateway. Use tracesGateway.config.merge, which will merge provided configuration to default configuration or tracesGateway.config.override which will override the whole configuration.
• For tracesSampler. Use tracesSampler.config.merge, which will merge provided configuration to default configuration or tracesSampler.config.override, which will override the whole configuration.

Pointing tracing clients (instrumentation exporters) to the agent collectors

Using OTLP HTTP is recommended:

• OTLP HTTP: <RELEASE_NAME>-sumologic-otelagent.<NAMESPACE>:4318

Alternatively, if required, you can use other supported formats as well:

• Jaeger GRPC: <RELEASE_NAME>-sumologic-otelagent.<NAMESPACE>:14250
• Jaeger Thrift HTTP: <RELEASE_NAME>-sumologic-otelagent.<NAMESPACE>:14268
• Jaeger Thrift Compact (UDP): <RELEASE_NAME>-sumologic-otelagent.<NAMESPACE>:6831
• Zipkin: <RELEASE_NAME>-sumologic-otelagent.<NAMESPACE>:9411/api/v2/spans
• OTLP gRPC: <RELEASE_NAME>-sumologic-otelagent.<NAMESPACE>:4317
• OTLP HTTP/deprecated: <RELEASE_NAME>-sumologic-otelagent.<NAMESPACE>:55681

For example, when the default release name (collection) and namespace (sumologic) is used, the endpoints are following:

• OTLP HTTP: collection-sumologic-otelagent.sumologic:4318
• Jaeger GRPC: collection-sumologic-otelagent.sumologic:14250
• Jaeger Thrift HTTP: collection-sumologic-otelagent.sumologic:14268
• Jaeger Thrift Compact (UDP): collection-sumologic-otelagent.sumologic:6831
• Zipkin: collection-sumologic-otelagent.sumologic:9411/api/v2/spans
• OTLP gRPC: collection-sumologic-otelagent.sumologic:4317
• OTLP HTTP/deprecated: collection-sumologic-otelagent.sumologic:55681

Troubleshooting

Desired Kubernetes installation state

After enabling and installing tracing, you should have additional Kubernetes resources:

• otelcol-instrumentation - collector responsible for data collection and tagging
  • StatefulSet: collection-sumologic-otelcol-instrumentation
  • Pod: collection-sumologic-otelcol-instrumentation-<hash>
  • Service: collection-sumologic-otelagent
  • Service (deprecated): collection-sumologic-otelcol
  • Config Map: collection-sumologic-otelcol-instrumentation
• traces-gateway - collector responsible for traces load balancing
  • Deployment: collection-sumologic-traces-gateway
  • Pod: collection-sumologic-traces-gateway-<hash>-<hash>
  • Replica Set: collection-sumologic-traces-gateway-<hash>
  • Service: collection-sumologic-traces-gateway
  • Config Map: collection-sumologic-traces-gateway
• traces-sampler - collector responsible for forwarding data to Sumo Receiver
  • Deployment: collection-sumologic-traces-sampler
  • Pod: collection-sumologic-traces-sampler-<hash>-<hash>
  • Replica Set: collection-sumologic-traces-sampler-<hash>
  • Service: collection-sumologic-traces-sampler-headless
  • Config Map: collection-sumologic-traces-sampler

How to verify traces are installed and working

• There are no Kubernetes errors in the namespace sumologic.

• There are running pods <RELEASE_NAME>-sumologic-otelcol-instrumentation-<hash>, <RELEASE_NAME>-sumologic-traces-gateway-<hash>, <RELEASE_NAME>-sumologic-traces-sampler-<hash>

• Kubernetes metadata tags such as pod and replicaset should be applied to all spans.

• The OpenTelemetry Collector can export metrics, which include information such as the number of spans exported. Several metrics starting with otelcol_ will become available, such as otelcol_exporter_sent_spans and otelcol_receiver_accepted_spans.

• OpenTelemetry Collector can have Debug exporter enabled. This will put on the output contents of spans (with some sampling above a certain rate). To enable, apply the following flags when installing/upgrading the collector (appending debug to the list of exporters):

  helm upgrade collection sumologic/sumologic \
    --namespace sumologic \
    ...
    --set tracesSampler.config.exporters.debug.verbosity=detailed \
    --set tracesSampler.config.service.pipelines.traces.exporters="{otlphttp,debug}"

  Having this enabled, kubectl logs -n sumologic collection-sumologic-traces-sampler-<ENTER ACTUAL POD ID> might yield the following output:

      2020-03-09T10:47:28.861Z TraceData with 1 spans
      Node service name: carpogonial
      Node attributes:
      2020-03-09T10:47:28.861Z Span #0
        Trace ID    : 00000000000000004abaf4a8688cee33
        ID          : 1aad0bc2b44e8219
        Parent ID   :
        Name        : Carpoidea
        Kind        : CLIENT
        Start time  : seconds:1583750845 nanos:799855000
        End time    : seconds:1583751016 nanos:332705000
        Span attributes:
              -> zipkin.remoteEndpoint.ipv6: 5ab8:31e6:a7b:6205:13cb:a3fe:c180:ca26
              -> ip: 10.1.1.1
              -> zipkin.remoteEndpoint.port: 49088
              -> zipkin.remoteEndpoint.serviceName: carpogonial
              -> ipv4: 36.110.13.238
              -> ipv6: 5ab8:31e6:a7b:6205:13cb:a3fe:c180:ca26
              -> port: 49088
              -> zipkin.remoteEndpoint.ipv4: 36.110.13.238