Skip to content

InfluxDB Exporter#
Description InfluxDB Exporter exports alarms, resources, or query results to InfluxDB.
Author Nokia
Catalog nokia-eda/catalog
Language Go

Overview#

The InfluxDB Exporter app exports Nokia EDA data into an InfluxDB server. It provides two groups of resources:

  • Server and ClusterServer: define the InfluxDB destination and connection settings
  • Export and ClusterExport: define what data to write and which server and bucket should receive it

The current app API is influxdb.eda.nokia.com/v1alpha1.

Installation#

The InfluxDB Exporter app can be installed using Nokia EDA Store or by running an AppInstaller workflow with kubectl or edactl:

apiVersion: appstore.eda.nokia.com/v1
kind: AppInstaller
metadata:
  name: influxdb-exporter-install
  namespace: eda-system
spec:
  operation: install
  apps:
    - appId: influxdb.eda.nokia.com
      catalog: eda-catalog-builtin-apps
      version:
        type: alias
        value: latest

cat << 'EOF' | kubectl apply -f -
apiVersion: appstore.eda.nokia.com/v1
kind: AppInstaller
metadata:
  name: influxdb-exporter-install
  namespace: eda-system
spec:
  operation: install
  apps:
    - appId: influxdb.eda.nokia.com
      catalog: eda-catalog-builtin-apps
      version:
        type: alias
        value: latest

EOF

Install Settings#

The app supports install-time settings through spec.apps[].appSettings in the AppInstaller workflow spec:

  • influxDBExporterCPULimit: CPU limit for the exporter pod. Default: "1"
  • influxDBExporterMemoryLimit: memory limit for the exporter pod. Default: "1Gi"
  • influxDBProxyConfig: ConfigMap name used for HTTP_PROXY, HTTPS_PROXY, and NO_PROXY. Default: proxy-config

The default requests are set to 500m CPU and 500Mi memory.

These settings control the deployment in the Nokia EDA base namespace and can be provided through spec.apps[].appSettings in the AppInstaller workflow or directly in the Nokia EDA UI.

apiVersion: appstore.eda.nokia.com/v1
kind: AppInstaller
metadata:
  name: influxdb-exporter-install-sized
  namespace: eda-system
spec:
  operation: install
  apps:
    - appId: influxdb.eda.nokia.com
      catalog: eda-catalog-builtin-apps
      version:
        type: alias
        value: latest
      appSettings:
        influxDBExporterCPULimit: "2"
        influxDBExporterMemoryLimit: 2Gi
        influxDBProxyConfig: proxy-config

cat << 'EOF' | kubectl apply -f -
apiVersion: appstore.eda.nokia.com/v1
kind: AppInstaller
metadata:
  name: influxdb-exporter-install-sized
  namespace: eda-system
spec:
  operation: install
  apps:
    - appId: influxdb.eda.nokia.com
      catalog: eda-catalog-builtin-apps
      version:
        type: alias
        value: latest
      appSettings:
        influxDBExporterCPULimit: "2"
        influxDBExporterMemoryLimit: 2Gi
        influxDBProxyConfig: proxy-config

EOF

Getting Started#

After installation, create:

  • a Server or ClusterServer to define how to connect to InfluxDB instance
  • an Export or ClusterExport to define what data should be exported and which InfluxDB bucket to write to

Namespace rules:

  • Server and Export are used within a user namespace such as eda
  • ClusterServer and ClusterExport are used from the EDA base namespace for centralized exports

Server Resources#

A server resource defines the target InfluxDB connection.

Notable specification fields:

  • url: InfluxDB base URL
  • org: InfluxDB organization
  • credentialsSecret: Secret containing either token or both username and password
  • batchSize: number of points buffered per write request
  • flushTimer: flush interval for buffered writes
  • timestampPrecision: seconds, milliseconds, microseconds, or nanoseconds
  • useGzip: enable gzip compression on write requests
  • tls: optional TLS configuration

TLS options:

  • tls.fromFiles: read caFile, certFile, and keyFile from mounted files
  • tls.fromSecret.name: read tls.crt, tls.key, and optional ca.crt from a Secret in the same namespace as the server resource

The resource status contains the fields related to the connectivity parameters of this instance:

  • connected
  • error
  • lastChecked
Example Server resource
apiVersion: influxdb.eda.nokia.com/v1alpha1
kind: Server
metadata:
  name: server1
  namespace: eda
spec:
  url: http://influxdb.example.svc.cluster.local:8086
  org: eda_team
  credentialsSecret: influx-secret
  batchSize: 100
  flushTimer: 10s
  timestampPrecision: milliseconds
  useGzip: false

cat << 'EOF' | kubectl apply -f -
apiVersion: influxdb.eda.nokia.com/v1alpha1
kind: Server
metadata:
  name: server1
  namespace: eda
spec:
  url: http://influxdb.example.svc.cluster.local:8086
  org: eda_team
  credentialsSecret: influx-secret
  batchSize: 100
  flushTimer: 10s
  timestampPrecision: milliseconds
  useGzip: false

EOF

Export Resources#

An export resource defines what data is written and which server and bucket receive it.

Each spec.servers[] entry references:

  • name: the Server or ClusterServer resource
  • bucket: the InfluxDB bucket to write to

Supported sources:

  • spec.exports.alarms
  • spec.exports.resource
  • spec.exports.query

Alarm Source#

Use spec.exports.alarms to export Nokia EDA alarms.

Important fields:

  • include: list of alarm types to include, or ["*"] for all alarms
  • exclude: optional list of alarm types to suppress

For ClusterExport, spec.exports.alarms.namespaces optionally limits which namespaces contribute alarms.

Alarm points are written with:

  • measurement name alarms
  • tags such as group, kind, type, resource, severity, sourceGroup, sourceKind, and sourceResource
  • remaining alarm fields stored as Influx fields

Resource Source#

Use spec.exports.resource[] to stream resource data from Nokia EDA resources.

Important fields:

  • group, version, kind: identify the CR type to export
  • name: optional single resource name
  • namespaces: cluster export only; limit which namespaces contribute resources

Resource points are written with:

  • measurement name equal to the lowercased resource kind
  • tags built from JsPath keys and Kubernetes labels
  • flattened object content as fields

Query Source#

Use spec.exports.query[] to write arbitrary state DB query results.

Important fields:

  • path: JsPath to export
  • fields: optional list of fields; if omitted, all fields are exported
  • where: optional EQL filter
  • mode: on-change, periodic, or both
  • period: required for periodic behavior
  • customization: optional regex-based measurement, tag, and field renaming

Path behavior:

  • for Export, omit the .namespace... prefix because the exporter automatically scopes the path to the export namespace
  • for ClusterExport, use the full path, typically including .namespace... when you want cross-namespace data

Query points are written with:

  • measurement name derived from the JsPath, joined with underscores
  • tags derived from JsPath keys
  • flattened object content as fields
Example Export resource
apiVersion: influxdb.eda.nokia.com/v1alpha1
kind: Export
metadata:
  name: interface-query
  namespace: eda
spec:
  description: Export enabled interface state to InfluxDB
  enabled: true
  exports:
    query:
      - path: .node.srl.interface
        fields:
          - admin-state
          - oper-state
          - name
        where: 'admin-state = "enable"'
        mode: periodic
        period: 10s
        customization:
          measurement:
            match: namespace_node_srl_interface
            replacement: interface_query
          fields:
            - match: ^admin-state$
              replacement: state
  servers:
    - name: server1
      bucket: eda

cat << 'EOF' | kubectl apply -f -
apiVersion: influxdb.eda.nokia.com/v1alpha1
kind: Export
metadata:
  name: interface-query
  namespace: eda
spec:
  description: Export enabled interface state to InfluxDB
  enabled: true
  exports:
    query:
      - path: .node.srl.interface
        fields:
          - admin-state
          - oper-state
          - name
        where: 'admin-state = "enable"'
        mode: periodic
        period: 10s
        customization:
          measurement:
            match: namespace_node_srl_interface
            replacement: interface_query
          fields:
            - match: ^admin-state$
              replacement: state
  servers:
    - name: server1
      bucket: eda

EOF

Cluster-Scoped Resources#

Use ClusterServer and ClusterExport when you want centralized writes from the Nokia EDA base namespace.

Cluster-specific behavior:

  • ClusterExport supports namespace filters for alarm and resource sources
  • ClusterExport can query across namespaces
  • ClusterServer uses the same connection model as Server, but is intended for centralized use
Example ClusterServer resource
apiVersion: influxdb.eda.nokia.com/v1alpha1
kind: ClusterServer
metadata:
  name: cluster-server1
  namespace: eda-system
spec:
  url: http://influxdb.eda-system.svc.cluster.local:8086
  org: eda_team
  credentialsSecret: influx-secret
  batchSize: 100
  flushTimer: 10s
  timestampPrecision: milliseconds
  useGzip: false

cat << 'EOF' | kubectl apply -f -
apiVersion: influxdb.eda.nokia.com/v1alpha1
kind: ClusterServer
metadata:
  name: cluster-server1
  namespace: eda-system
spec:
  url: http://influxdb.eda-system.svc.cluster.local:8086
  org: eda_team
  credentialsSecret: influx-secret
  batchSize: 100
  flushTimer: 10s
  timestampPrecision: milliseconds
  useGzip: false

EOF
Example ClusterExport resource
/// tab | YAML

```yaml
apiVersion: influxdb.eda.nokia.com/v1alpha1
kind: ClusterExport
metadata:
  name: cluster-alarms
  namespace: eda-system
spec:
  description: Export alarms from selected namespaces to InfluxDB
  enabled: true
  exports:
    alarms:
      namespaces:
        - eda
        - tenant-a
      include:
        - "*"
      exclude:
        - InfluxDBServerConnectionFailed
  servers:
    - name: cluster-server1
      bucket: eda

cat << 'EOF' | kubectl apply -f -
apiVersion: influxdb.eda.nokia.com/v1alpha1
kind: ClusterExport
metadata:
  name: cluster-alarms
  namespace: eda-system
spec:
  description: Export alarms from selected namespaces to InfluxDB
  enabled: true
  exports:
    alarms:
      namespaces:
        - eda
        - tenant-a
      include:
        - "*"
      exclude:
        - InfluxDBServerConnectionFailed
  servers:
    - name: cluster-server1
      bucket: eda

EOF

Connectivity Alarm#

The app raises an EDA alarm when it cannot connect to a configured InfluxDB server.

Alarm details:

  • Alarm type: InfluxDBServerConnectionFailed
  • Severity: major
  • Resource kind: Server or ClusterServer
  • Resource: server name

Typical causes:

  • wrong InfluxDB URL or port
  • invalid token or username/password secret
  • bad TLS certificate, key, or CA configuration
  • network connectivity or proxy configuration issues

The alarm clears automatically after successful connectivity checks.

Validation Notes#

When creating resources, follow these rules:

  • flushTimer must be at least 1s
  • batchSize must be at least 1
  • if a TLS client certificate or key is set in tls.fromFiles, the matching key or certificate must also be set
  • every export must define at least one source and at least one server
  • alarm exports must set include or exclude
  • resource exports must set group, version, and kind
  • query customizations use regexes, so invalid regex patterns are rejected