Aspecto.io opentelemetry sampling collector
Welcome to the helm chart of the aspecto.io opentelemetry sampling collector.
The collector allows you to install traces collector on your kubernetes cluster and define rules of when the traces should be dropped and when they should be sent to your aspecto
- Kubernetes 1.12+
- Helm 3.1.0 or higher
- Aspecto.io account and api token (which will be mapped in the environment variables of the collector)
(*) we encourage you to use a DNS CNAME, it will ease on you possible changes in your deployment configuration with smaller footprint on production.
- add the aspecto.io help repo by running:
helm repo add aspecto https://aspecto-io.github.io/helm-charts- go to https://app.aspecto.io/app/integration/api-key and copy your api key
- run the install command:
helm install --name aspecto-collector aspecto/aspecto-io-opentelemetry-collector --set global.aspecto.token.value="<token>"you can change the default values stored in the values file according to the specified configuration list bellow. you can do that using the following options
- pass new value using the set command
helm install --set <key>=<value>- change the content of the
values.yamlfile and run the helm install command without the set part. - create additional values file (e.g. production.yaml) and install the chart using the following command
helm install --name aspecto-collector -f values.yaml -f production.yamlIn order to change configuration or upgrade please run the following command, the same aspects of set / file configuration applies here.
helm upgrade --name aspecto-collector -f values.yaml -f production.yamlThis section of parameters will be shared across all objects unless specified per service
| Name | Description | Type/Options | Default | Required |
|---|---|---|---|---|
global.name |
Global prefix for objects (deployment/service/etc.) | string | empty | true |
global.env |
Global environment variables | object | {} | false |
Global.aspecto : aspecto configuration |
||||
global.aspecto.interval |
refresh rate to pull sampling configuration (in seconds) | number | 60 | true |
Global.rolloutStrategy : deployment rollout configuration |
||||
global.rolloutStrategy.type |
rolling deployment strategy | Recreate , RollingUpdate |
RollingUpdate | true |
global.rolloutStrategy.rollingUpdate.maxUnavailable |
How many instances are allowed to be unavailable during the deployment | number | 0 | true |
global.rolloutStrategy.rollingUpdate.maxSurge |
How many instances should be spawned up before taking down existing instances | number | 2 | true |
Global.metadata : objects metada configuration |
||||
global.metadata.shared.annotations |
Global annotations to be shared across all child objects | object | {} | false |
global.metadata.shared.labels |
Global labels to be shared across all child objects | object | {} | false |
global.metadata.{pod/deployment/service/configMap/hpa/virtualService/destinationRule}.annotations |
Global annotations to be shared across all pods | object | {} | false |
global.metadata.{pod/deployment/service/configMap/hpa/virtualService/destinationRule}.labels |
Global labels to be shared across all selected object | object | {} | false |
Global.image : global image configuration (will be overridden in case specified in service) |
||||
global.image.repository |
Global repository to pull the image from | string | public.ecr.aws/x3s3n8k7 | false |
global.image.pullPolicy |
Global repository to pull the image from | Always, Never, IfNotPresent |
Always | false |
global.image.version |
Global image to pull (we're using the same tag when new version is release and tested) | string | specified to latest stable version | false |
This section of parameters describes the receiver service
| Name | Description | Type/Options | Default | Required |
|---|---|---|---|---|
receiver.env |
environment variables | object | {} | false |
receiver.image |
_This section of parameters will be configure image policy. leaving values empty will populate values from the global section | |||
receiver.image.repository |
repository to pull the image from | string | public.ecr.aws/x3s3n8k7 | false |
receiver.image.pullPolicy |
repository to pull the image from | Always, Never, IfNotPresent |
Always | false |
receiver.image.version |
image to pull (we're using the same tag when new version is release and tested) | string | specified to latest stable version | false |
receiver.image.name |
load balancing docker image name | string | otelcol-loadbalancing | true |
receiver.metadata |
||||
receiver.metadata.shared.annotations |
annotations to be shared across all child objects | object | {} | false |
receiver.metadata.shared.labels |
labels to be shared across all child objects | object | {} | false |
receiver.metadata.{pod/deployment/service/configMap/hpa/virtualService/destinationRule}.annotations |
annotations to be shared across all pods | object | {} | false |
receiver.metadata.{pod/deployment/service/configMap/hpa/virtualService/destinationRule}.labels |
labels to be shared across all selected object | object | {} | false |
receiver.envs : This section of parameters define environment variables. |
||||
| receiver.envs | environment variables that will be defined on the deployment pods. value can be defined from any source (static, value, etc.) | array | [] | true |
| receiver.envsFrom | environment variables that which will be imported and defined on the deployment pods. | array | [] | true |
receiver.specs : This section of parameters define resources over the service |
||||
receiver.specs.configuration : service configuration |
||||
receiver.specs.configuration.exporters.load_balancing.protocol.otlp.timeout |
timeout of connection in case of no response | string | 1s | true |
receiver.specs.configuration.exporters.load_balancing.protocol.otlp.tls.insecure |
connection over TLS should verify certificate | boolean | false | true |
receiver.specs.configuration.exporters.load_balancing.protocol.resolver.dns.hostname |
sampling service endpoint | string | otel-collector-opentelemetry-sampling-service | false |
receiver.specs.configuration.exporters.load_balancing.protocol.resolver.dns.port |
sampling service grpc port number | number | 4317 | true |
receiver.specs.configuration.log.enable |
should turn on logger | boolean | false | false |
receiver.specs.configuration.log.level |
log level | fatal, error, info, debug |
error | false |
receiver.specs.configuration.endpoint |
endpoint for virtual service | string | '' | false |
receiver.specs.autoscaling : hpa/autoscaling configuration |
||||
receiver.specs.autoscaling.enable |
Turn on autoscaling | boolean | true | true |
receiver.specs.autoscaling.defaultReplicaCount |
replicaCount in case not using HPA | number | 3 | true |
receiver.specs.autoscaling.minReplicas |
minimum amount of pods | number | 3 | false |
receiver.specs.autoscaling.maxReplicas |
maximum amount of pods | number | 20 | false |
receiver.specs.autoscaling.targetCPUUtilizationPercentage |
CPU consumption percentage threshold before creating more pods | number | 75 | false |
receiver.specs.autoscaling.targetMemoryUtilizationPercentage |
Memory consumption percentage threshold before creating more pods | number | 75 | false |
receiver.specs.resources : resources configuration |
||||
receiver.specs.resources..limits.cpu |
Amount of cores to allocate to the pod (hard limit) | number | 2 | true |
receiver.specs.resources.limits.memory |
Amount of memory to allocate to the pod (hard limit) | string | 2Gi | true |
receiver.specs.resources.requests.cpu |
Amount of cores to allocate to the pod (soft limit) | number | 1 | true |
receiver.specs.resources.requests.memory |
Amount of cores to allocate to the pod (soft limit) | string | 1Gi | true |
receiver.specs.probe : pod state monitoring |
||||
receiver.specs.probe.readinessProbe.httpGet.path |
(Readiness) HTTP uri to run health-check against | string | / | true |
receiver.specs.probe.readinessProbe.httpGet.port |
(Readiness) HTTP port to run health-check against | number | 8090 | true |
receiver.specs.probe.readinessProbe.initialDelaySeconds |
(Readiness) How many seconds to wait before first request to probe the state | number | 10 | true |
receiver.specs.probe.readinessProbe.periodSeconds |
(Readiness) TTL of request to probe the state | number | 3 | true |
receiver.specs.probe.livenessProbe.httpGet.path |
(Liveness) HTTP uri to run health-check against | string | / | true |
receiver.specs.probe.livenessProbe.httpGet.port |
(Liveness) HTTP port to run health-check against | number | 8090 | true |
receiver.specs.probe.livenessProbe.initialDelaySeconds |
(Liveness) How many seconds to wait before first request to probe the state | number | 10 | true |
receiver.specs.probe.livenessProbe.periodSeconds |
(Liveness) TTL of request to probe the state | number | 3 | true |
receiver.specs.volumes : pod volumes configuration |
||||
receiver.specs.volumes.serviceConfig.name |
Volume name | string | otel-lb-config | true |
receiver.specs.volumes.serviceConfig.mountPath |
mount folder location (name) | string | /config | true |
receiver.specs.volumes.serviceConfig.fileName |
configuration file name | string | config.yaml | true |
receiver.specs.volumes.extraVolumes |
array of additional volumes | [] | [] | false |
receiver.specs.volumes.extraVolumes.name |
volume name | string | false | |
receiver.specs.volumes.extraVolumes.mountPath |
mount folder path | string | false | |
receiver.specs.volumes.extraVolumes.subPath |
filename | string | false | |
receiver.specs.volumes.extraSecretMounts |
array of additional secrets volumes | [] | [] | false |
receiver.specs.volumes.extraSecretMounts.[].name |
volume name | string | false | |
receiver.specs.volumes.extraSecretMounts.[].mountPath |
file location in the pod | string | false | |
receiver.specs.volumes.extraSecretMounts.[].secretName |
secret name | string | false | |
receiver.specs.volumes.extraSecretMounts.[].readOnly |
should use readOnly mount | boolean | true | false |
receiver.specs.network : pod network configuration |
default configuration already applied | |||
receiver.specs.network.localListenerIp |
IP address the application should accept requests from | string | 0.0.0.0 | true |
receiver.specs.network.type |
Network type | string | None | true |
receiver.specs.network.ports |
Port configuration array | [] | [] | true |
receiver.specs.network.ports.[].name |
Port name | string | grpc | true |
receiver.specs.network.ports.[].internalPort |
egress port number | number | 4317 | true |
receiver.specs.network.ports.[].externalPort |
ingress port number | number | 4317 | true |
receiver.specs.network.ports.[].protocol |
Port protocol | TCP, UDP, ICMP |
TCP | true |
receiver.specs.network.istio.enable |
Should enable istio | boolean | false | true |
receiver.specs.network.istio.rules |
istio rules | [] | [] | false |
receiver.specs.services : service configuration |
||||
receiver.specs.service.[].name |
Service name | string | opentelemetry-sampling-receiver | true |
receiver.specs.service.[].serviceType |
Service type | LoadBalancer, ClusterIP |
LoadBalancer | true |
receiver.specs.service.[].selector |
key:value of pod labels | {} | component: otel-receiver-pod | true |
receiver.specs.service.[].ports |
service exposed ports | [] | (default configuration provided) | true |
receiver.specs.service.[].ports.[].name |
port name | string | grpc | true |
receiver.specs.service.[].ports.[].type |
port type | string | grpc | true |
receiver.specs.service.[].ports.[].externalPort |
external port number | number | 4318 | true |
receiver.specs.service.[].ports.[].internalPort |
internal port number | number | 4318 | true |
receiver.specs.service.[].ports.[].protocol |
port protocol | TCP, UDP, ICMP |
TCP | false |
receiver.specs.service.[].ports.[].loadBalancer |
routing mechanism | object | simple: LEAST_CONN | true |
This section of parameters describes the collector service
| Name | Description | Type/Options | Default | Required |
|---|---|---|---|---|
collector.env |
environment variables | object | {} | false |
collector.image |
_This section of parameters will be configure image policy. leaving values empty will populate values from the global section | |||
collector.image.repository |
repository to pull the image from | string | public.ecr.aws/x3s3n8k7 | false |
collector.image.pullPolicy |
repository to pull the image from | Always, Never, IfNotPresent |
Always | false |
collector.image.version |
image to pull (we're using the same tag when new version is release and tested) | string | specified to latest stable version | false |
collector.image.name |
load balancing docker image name | string | otelcol-sampling | true |
collector.metadata |
||||
collector.metadata.shared.annotations |
annotations to be shared across all child objects | object | {} | false |
collector.metadata.shared.labels |
labels to be shared across all child objects | object | {} | false |
collector.metadata.{pod/deployment/service/configMap/hpa/virtualService/destinationRule}.annotations |
annotations to be shared across all pods | object | {} | false |
collector.metadata.{pod/deployment/service/configMap/hpa/virtualService/destinationRule}.labels |
labels to be shared across all selected object | object | {} | false |
collector.envs : This section of parameters define environment variables. |
||||
| collector.envs | environment variables that will be defined on the deployment pods. one of them must be the token variable. value can be defined from any source (static, value, etc.) | array | [] | true |
| collector.envsFrom | environment variables that which will be imported and defined on the deployment pods. | array | [] | true |
collector.specs : This section of parameters define resources over the service |
||||
collector.specs.configuration : service configuration |
||||
collector.specs.configuration.endpoint |
timeout of connection in case of no response | string | 1s | true |
collector.specs.configuration.decision_wait |
number of seconds to wait before closing the trace | string | 10s | true |
collector.specs.configuration.num_traces |
number of traces to hold in the memory | number | 100 | true |
collector.specs.configuration.expected_new_traces_per_sec |
number of expected traces per seconds | number | 10 | true |
collector.specs.configuration.policies |
policy name (aspecto mapping | string | $$aspecto:tail_sampling_policy | true |
collector.specs.configuration.log.enable |
should enable logging | boolean | false | false |
collector.specs.configuration.log.level |
log level | fatal, error, info, debug |
error | false |
collector.specs.autoscaling : hpa/autoscaling configuration |
||||
collector.specs.autoscaling.enable |
Turn on autoscaling | boolean | true | true |
collector.specs.autoscaling.defaultReplicaCount |
replicaCount in case not using HPA | number | 3 | true |
collector.specs.autoscaling.minReplicas |
minimum amount of pods | number | 3 | false |
collector.specs.autoscaling.maxReplicas |
maximum amount of pods | number | 20 | false |
collector.specs.autoscaling.targetCPUUtilizationPercentage |
CPU consumption percentage threshold before creating more pods | number | 75 | false |
collector.specs.autoscaling.targetMemoryUtilizationPercentage |
Memory consumption percentage threshold before creating more pods | number | 75 | false |
collector.specs.resources : resources configuration |
||||
collector.specs.resources..limits.cpu |
Amount of cores to allocate to the pod (hard limit) | number | 2 | true |
collector.specs.resources.limits.memory |
Amount of memory to allocate to the pod (hard limit) | string | 2Gi | true |
collector.specs.resources.requests.cpu |
Amount of cores to allocate to the pod (soft limit) | number | 1 | true |
collector.specs.resources.requests.memory |
Amount of cores to allocate to the pod (soft limit) | string | 1Gi | true |
collector.specs.probe : pod state monitoring |
||||
collector.specs.probe.readinessProbe.httpGet.path |
(Readiness) HTTP uri to run health-check against | string | / | true |
collector.specs.probe.readinessProbe.httpGet.port |
(Readiness) HTTP port to run health-check against | number | 8090 | true |
collector.specs.probe.readinessProbe.initialDelaySeconds |
(Readiness) How many seconds to wait before first request to probe the state | number | 10 | true |
collector.specs.probe.readinessProbe.periodSeconds |
(Readiness) TTL of request to probe the state | number | 3 | true |
collector.specs.probe.livenessProbe.httpGet.path |
(Liveness) HTTP uri to run health-check against | string | / | true |
collector.specs.probe.livenessProbe.httpGet.port |
(Liveness) HTTP port to run health-check against | number | 8090 | true |
collector.specs.probe.livenessProbe.initialDelaySeconds |
(Liveness) How many seconds to wait before first request to probe the state | number | 10 | true |
collector.specs.probe.livenessProbe.periodSeconds |
(Liveness) TTL of request to probe the state | number | 3 | true |
collector.specs.volumes : pod volumes configuration |
||||
collector.specs.volumes.serviceConfig.name |
Volume name | string | otel-lb-config | true |
collector.specs.volumes.serviceConfig.mountPath |
mount folder location (name) | string | /config | true |
collector.specs.volumes.serviceConfig.fileName |
configuration file name | string | config.yaml | true |
collector.specs.volumes.extraVolumes |
array of additional volumes | [] | [] | false |
collector.specs.volumes.extraVolumes.name |
volume name | string | false | |
collector.specs.volumes.extraVolumes.mountPath |
mount folder path | string | false | |
collector.specs.volumes.extraVolumes.subPath |
filename | string | false | |
collector.specs.volumes.extraSecretMounts |
array of additional secrets volumes | [] | [] | false |
collector.specs.volumes.extraSecretMounts.[].name |
volume name | string | false | |
collector.specs.volumes.extraSecretMounts.[].mountPath |
file location in the pod | string | false | |
collector.specs.volumes.extraSecretMounts.[].secretName |
secret name | string | false | |
collector.specs.volumes.extraSecretMounts.[].readOnly |
should use readOnly mount | boolean | true | false |
collector.specs.network : pod network configuration |
default configuration already applied | |||
collector.specs.network.localListenerIp |
IP address the application should accept requests from | string | 0.0.0.0 | true |
collector.specs.network.type |
Network type | string | None | true |
collector.specs.network.ports |
Port configuration array | [] | [] | true |
collector.specs.network.ports.[].name |
Port name | string | grpc | true |
collector.specs.network.ports.[].internalPort |
egress port number | number | 4317 | true |
collector.specs.network.ports.[].externalPort |
ingress port number | number | 4317 | true |
collector.specs.network.ports.[].protocol |
Port protocol | TCP, UDP, ICMP |
TCP | true |
collector.specs.network.istio.enable |
Should enable istio | boolean | false | true |
collector.specs.network.istio.rules |
istio rules | [] | [] | false |
collector.specs.services : service configuration |
||||
collector.specs.service.[].name |
Service name | string | opentelemetry-sampling | true |
collector.specs.service.[].serviceType |
Service type | LoadBalancer, ClusterIP |
ClusterIP | true |
collector.specs.service.[].selector |
key:value of pod labels | {} | component: otel-receiver-pod | true |
collector.specs.service.[].ports |
service exposed ports | [] | (default configuration provided) | true |
collector.specs.service.[].ports.[].name |
port name | string | grpc | true |
collector.specs.service.[].ports.[].type |
port type | string | grpc | true |
collector.specs.service.[].ports.[].externalPort |
external port number | number | 8081 | true |
collector.specs.service.[].ports.[].internalPort |
internal port number | number | 4318 | true |
collector.specs.service.[].ports.[].protocol |
port protocol | TCP, UDP, ICMP |
TCP | false |
collector.specs.service.[].ports.[].loadBalancer |
routing mechanism | object | simple: LEAST_CONN | true |
This section of parameters describes the redis service
redis is used for taking linked traces sampling decisions across multiple collector instances
| Name | Description | Type/Options | Default | Required |
|---|---|---|---|---|
| redis.enabled | enable/disable linked traces sampling decisions | boolean | false | true |
redis.config |
This section of parameters will configure linked traces redis policy | |||
redis.config.type |
redis type, local means starting a redis deployment as part of this chart, external means to use an external redis (you should use external in cases of very high throughput) | local, external |
string | local |
redis.config.endpoint |
redis endpoint - only relevant in case of using external type |
string | '' | false |
redis.config.ttlMinutes |
redis item ttl in minutes - it means that the linked trace sampling decision need to take place before this ttl | number | 5 | false |
redis.config.credentials |
Configure redis username and password | |||
redis.config.credentials.username |
redis username | string | '' | false |
redis.config.credentials.password |
redis password | string | '' | false |
redis.config.credentials.secret |
In case we want to use known secrets | |||
redis.config.credentials.secret.name |
secret name | string | '' | false |
redis.config.credentials.secret.username_key |
secret username key | string | '' | false |
redis.config.credentials.secret.password_key |
secret name | string | '' | false |
redis.metadata |
objects metadata configuration | |||
redis.metadata.annotations |
annotations to be shared across all child objects | object | {} | false |
redis.metadata.labels |
labels to be shared across all child objects | object | {} | false |
redis.spec |
This section of parameters define resources over the service | |||
redis.spec.replicas |
deployment replica count | number | 1 | false |
redis.spec.port |
redis port | string | 6379 | false |
redis.spec.image |
This section of parameters define image resources | |||
redis.spec.image.policy |
Pull image policy | Always, Never, IfNotPresent |
Always | false |
redis.spec.image.name |
Image name | redis | false | |
redis.spec.image.version |
Image version | 6.0.8 | false | |
redis.spec.resources |
resources configuration | |||
redis.spec.resources.limits.cpu |
Amount of cores to allocate to the pod (hard limit) | number | 1 | false |
redis.spec.resources.limits.memory |
Amount of memory to allocate to the pod (hard limit) | string | 2Gi | false |
redis.spec.resources.requests.cpu |
Amount of cores to allocate to the pod (soft limit) | number | 0.5 | false |
redis.spec.resources.requests.memory |
Amount of cores to allocate to the pod (soft limit) | string | 2Gi | false |
We suggest following the general guidelines provided by the otel community, as listed below. You are always welcome to re-check for the latest stress-test results.
| Span Format |
CPU (2+ GHz) |
RAM (GB) |
Sustained Rate |
Recommended Maximum |
|---|---|---|---|---|
| OpenTelemetry | 1 | 2 | ~9K | 8K |
| OpenTelemetry | 2 | 4 | ~18K | 16K |