influxdata / helm-charts

Official Helm Chart Repository for InfluxData Applications

InfluxData Helm charts

Usage

1. Install Helm. For more information, see Helm documentation.

2. Add the InfluxData Helm repository:

   helm repo add influxdata https://helm.influxdata.com/

3. View InfluxData Helm charts:

   helm search repo influxdata

Support

These Helm charts are community supported. InfluxData provides no official support for their use.

Pull requests and issues are the responsibility of the project's moderator(s) which may include vetted individuals outside of the InfluxData organization. All issues should be reported and managed via GitHub (not via InfluxData's standard support process).

Join us on Slack in #helm!

Contributing

We'd love to have you contribute! Please refer to our contribution guidelines for details.

License

MIT License

helm-charts's Issues

Update Chronograf Helm chart to support connecting to etcd backend

Chronograf 1.8 adds the ability to use etcd as the backing store with the following config options:
https://docs.influxdata.com/chronograf/v1.8/administration/config-options/#etcd-options

We should add those to the helm chart

[telegraf-operator] Default install does not work

Hi,

Trying to discover telegraf-operator, default chart does not work:

kubectl create ns telegraf-operator
helm upgrade -i telegraf-operator -n telegraf-operator influxdata/telegraf-operator

It leads to:

NAME                                     READY   STATUS              RESTARTS   AGE
pod/telegraf-operator-55c9788d8b-lpltn   0/1     ContainerCreating   0          6m35s
pod/telegraf-operator-55c9788d8b-ppw5v   0/1     ContainerCreating   0          6m35s
pod/telegraf-operator-55c9788d8b-wq8zp   0/1     ContainerCreating   0          6m35s

NAME                        TYPE        CLUSTER-IP     EXTERNAL-IP   PORT(S)   AGE
service/telegraf-operator   ClusterIP   10.3.206.109   <none>        443/TCP   6m35s

NAME                                READY   UP-TO-DATE   AVAILABLE   AGE
deployment.apps/telegraf-operator   0/3     3            0           6m36s

NAME                                           DESIRED   CURRENT   READY   AGE
replicaset.apps/telegraf-operator-55c9788d8b   3         3         0       6m36s

Going further:

kubectl describe pod/telegraf-operator-55c9788d8b-lpltn                                                                                                    
Name:           telegraf-operator-55c9788d8b-lpltn
Namespace:      telegraf-operator
Priority:       0
Node:           node-43ff5d12-3a64-432d-b16d-7dd781ad101e/51.75.199.67
Start Time:     Thu, 14 May 2020 11:14:44 +0200
Labels:         app.kubernetes.io/instance=telegraf-operator
                app.kubernetes.io/name=telegraf-operator
                pod-template-hash=55c9788d8b
Annotations:    <none>
Status:         Pending
IP:             
IPs:            <none>
Controlled By:  ReplicaSet/telegraf-operator-55c9788d8b
Containers:
  telegraf-operator:
    Container ID:  
    Image:         quay.io/influxdb/telegraf-operator:v1.0.6
    Image ID:      
    Port:          9443/TCP
    Host Port:     0/TCP
    Args:
      --telegraf-default-class=infra
      --telegraf-classes-directory=/etc/telegraf-operator
      --enable-default-internal-plugin
      --telegraf-image=docker.io/library/telegraf:1.14.1
    State:          Waiting
      Reason:       ContainerCreating
    Ready:          False
    Restart Count:  0
    Limits:
      cpu:     100m
      memory:  128Mi
    Requests:
      cpu:     50m
      memory:  64Mi
    Environment:
      POD_NAMESPACE:  telegraf-operator (v1:metadata.namespace)
    Mounts:
      /etc/certs from certs (ro)
      /etc/telegraf-operator from classes (ro)
      /var/run/secrets/kubernetes.io/serviceaccount from telegraf-operator-token-ccw57 (ro)
Conditions:
  Type              Status
  Initialized       True 
  Ready             False 
  ContainersReady   False 
  PodScheduled      True 
Volumes:
  certs:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  telegraf-operator-tls
    Optional:    false
  classes:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  telegraf-operator-classes
    Optional:    false
  telegraf-operator-token-ccw57:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  telegraf-operator-token-ccw57
    Optional:    false
QoS Class:       Burstable
Node-Selectors:  <none>
Tolerations:     node.kubernetes.io/not-ready:NoExecute for 300s
                 node.kubernetes.io/unreachable:NoExecute for 300s
Events:
  Type     Reason       Age                 From                                                Message
  ----     ------       ----                ----                                                -------
  Normal   Scheduled    <unknown>           default-scheduler                                   Successfully assigned telegraf-operator/telegraf-operator-55c9788d8b-lpltn to node-43ff5d12-3a64-432d-b16d-7dd781ad101e
  Warning  FailedMount  13s                 kubelet, node-43ff5d12-3a64-432d-b16d-7dd781ad101e  Unable to attach or mount volumes: unmounted volumes=[classes], unattached volumes=[certs classes telegraf-operator-token-ccw57]: timed out waiting for the condition
  Warning  FailedMount  8s (x9 over 2m16s)  kubelet, node-43ff5d12-3a64-432d-b16d-7dd781ad101e  MountVolume.SetUp failed for volume "classes" : secret "telegraf-operator-classes" not found

So the classe telegraf-operator-classes which seems to be a secret (why not a configmap by the way ?) is not present.

It's because by default data is empty in values.yml :

classes:
  secretName: "telegraf-operator-classes"
  default: "infra"
  data: {}

So it should be mentionned in the README that we have to provide values for data to make it work as expected.

Also on the README, why is it stated that cert-manager is required if value is false by default ? Integration with cert-manager should be better explained.

[influxdb] add note that users should configure ephemeral storage request in backup job

when using emptyDir (default)... we could configure storage requests to ensure backup is scheduled on a node capable of supplying ephemeral storage as it backs up to object store

cc @monotek @rawkode

Adopt Semantic Commits

[influxdb 2.0] changes needed for GA release

TBD

1.0.4 version failed Kubernetes API

I checked an error in Kubernetes input plugin when I deployed 1.0.4 version:
Error in plugin: https://10.200.0.35:10250/pods returned HTTP status 403 Forbidden

Commands to reproduce, my helm chart had 1.0.3 version:

$ helm upgrade -f ./config.yaml --force telegraf-ds influxdata/telegraf-ds
Release "telegraf-ds" has been upgraded. Happy Helming!
NAME: telegraf-ds
LAST DEPLOYED: Tue Apr  7 13:38:48 2020
NAMESPACE: logstreaming
STATUS: deployed
REVISION: 5
TEST SUITE: None
NOTES:
[..]


$ kubectl logs telegraf-ds-ggzhn
2020-04-07T11:38:58Z I! Starting Telegraf 1.14.0
2020-04-07T11:38:58Z I! Using config file: /etc/telegraf/telegraf.conf
2020-04-07T11:38:58Z I! Loaded inputs: disk kubernetes diskio kernel swap cpu mem processes system docker
2020-04-07T11:38:58Z I! Loaded aggregators: 
2020-04-07T11:38:58Z I! Loaded processors: 
2020-04-07T11:38:58Z I! Loaded outputs: influxdb
2020-04-07T11:38:58Z I! Tags enabled: host=aks-otherpurpose-28286210-vmss000000
2020-04-07T11:38:58Z I! [agent] Config: Interval:10s, Quiet:false, Hostname:"aks-otherpurpose-28286210-vmss000000", Flush Interval:10s
2020-04-07T11:39:00Z E! [inputs.kubernetes] Error in plugin: https://10.200.0.35:10250/pods returned HTTP status 403 Forbidden
2020-04-07T11:39:10Z E! [inputs.kubernetes] Error in plugin: https://10.200.0.35:10250/pods returned HTTP status 403 Forbidden

But If upgrade to 1.0.3 revision again , is OK:

helm upgrade -f ./config.yaml --force telegraf-ds influxdata/telegraf-ds --version 1.0.3
Release "telegraf-ds" has been upgraded. Happy Helming!
NAME: telegraf-ds
LAST DEPLOYED: Tue Apr  7 13:39:45 2020
NAMESPACE: logstreaming
STATUS: deployed
REVISION: 6
TEST SUITE: None
NOTES:
[....]


$ kubectl logs telegraf-ds-2l5k2
2020-04-07T11:39:58Z I! Starting Telegraf 1.13.4
2020-04-07T11:39:58Z I! Using config file: /etc/telegraf/telegraf.conf
2020-04-07T11:39:58Z I! Loaded inputs: diskio kernel processes swap disk docker mem system cpu kubernetes
2020-04-07T11:39:58Z I! Loaded aggregators: 
2020-04-07T11:39:58Z I! Loaded processors: 
2020-04-07T11:39:58Z I! Loaded outputs: influxdb
2020-04-07T11:39:58Z I! Tags enabled: host=aks-otherpurpose-28286210-vmss000000
2020-04-07T11:39:58Z I! [agent] Config: Interval:10s, Quiet:false, Hostname:"aks-otherpurpose-28286210-vmss000000", Flush Interval:10s

Nothing errors with 1.0.3

[influxdb-enterprise] Add support TLS log debug configuration

With TLS certs configured, the logs become very verbose, please add configuration to allow turning on/off the TLS debug log in data pod for [meta}data] pods, do the same)

[influxdb-enterprise] config reference error

Misalignment of how the values are expected to be defined and how they are referenced (also for meta, expected config and referenced in chart). Changing the references to
.Values.[meta|data].https.secret.key appears to have fix the issue.

using telegraf-ds and influx as subcharts

It would be nice to be able put together an application helm chart that installs telegraf and influx as subcharts (grafana also). In order to do this. simple change to telegraf to use tpl function on output config can be done to accomplish this:

values.yaml:

telegraf:
  enabled: false
  config:
    outputs:
    - influxdb:
        urls:
          - "http://{{ .Release.Name }}-influxdb.{{ .Release.Namespace }}.svc:8086"

configmap.yaml:
would need to pass full context to output template and use tpl for string values.
{{ template "outputs" . }}

and outputs would be changed as.

{{- $context := . -}}
{{- range $outputIdx, $configObject := .Values.config.outputs -}}
...
          {{- if eq $tp "string" }}
      {{ $key }} = {{ tpl $value $context | quote }}
...```

Adding input.internal

I'm trying to add the input internal to the telegraf config with the following configuration:

## Default values.yaml for Telegraf
## This is a YAML-formatted file.
## ref: https://hub.docker.com/r/library/telegraf/tags/
...

  inputs:
    - kubernetes:
        url: "https://$HOSTIP:10250"
        bearer_token: "/var/run/secrets/kubernetes.io/serviceaccount/token"
        insecure_skip_verify: true
    - docker: 
        endpoint: "unix:///var/run/docker.sock"
        timeout: "5s"
        perdevice: true
        total: false
        docker_label_exclude:
        - "annotation.kubernetes.io/*"
    - internal:

But it's not added in telegraf.conf, how can I add it?

[chronograf] Support loading custom dashboards

It doesn't seem to be possible to include a custom dashboard when installing the chronograf chart. The way I would do it if I was writing the Deployment myself is to create a ConfigMap with the file:

apiVersion: v1
kind: ConfigMap
metadata:
  name: chronograf-dashboard
data:
  custom-dashboard.json: |
    {
      "id": "custom-dashboar",
      "measurement": "apache",
      "app": "apache",
      ...
    }

Then mount it:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: chronograf
  labels:
    app: chronograf
spec:
  replicas: 1
  selector:
    matchLabels:
      app: chronograf
  template:
    metadata:
      labels:
        app: chronograf
    spec:
      containers:
        - name: chronograf
          image: chronograf:latest
          ports:
            - containerPort: 8888
          volumeMounts:
          - name: dashboard-volume
            mountPath: /usr/share/chronograf/resources
      volumes:
        - name: dashboard-volume
          configMap:
            name: chronograf-dashboard

I'm willing to work on this but waiting on my company to sign the CCLA. In the meantime is there another way to include a custom dashboard? Thanks.

[influxdb] conditionally add user/auth to backup script

cc @monotek @rawkode

if setDefaultUser.enabled: true for example

[influxdb] Should support secret and volume mounts.

Hi all,

I have a selfsigned certificate generated by cert-manager that I would like to use to configure HTTPS for Influxdb (between the ingress and the service). Unfortunately, I can't seem to find any options that support this. If such an avenue exists, could someone please point me to it?

Update README Badge with Action Name

This was changed in CI upgrades

telegraf optional secret(s)

Add an option for a secret (or more) to be loaded by the telegraf deployment.

Helm Chart(s) for Telegraf

We should provide one or more helm charts for installing Telegraf in a Kubernetes cluster. We should support a Kubernetes monitoring template as well as an Istio monitoring template.

[chronograf] mismatch of accessMode variable name

values.yaml declares a variable persistence.accessModes but the template file expects a variable named persistence.accessMode

Export and test Istio Template

[ ] SRE dashboards
[ ] istio bucket
[ ] errorRate task
[ ] latency task
[ ] error rate bucket
[ ] latency bucket
[ ] telegraf config(s)

[influxdb-enterprise] Add support for aggregating custom certs

Add option to allow aggregating custom CA cert on top of all the public certs vs letting custom CA cert overriding public certs (so only those trusted by custom CA is allowed)

[charts/influxdb] The influxdb chart needs service-less service.

Describe the bug:

https://kubernetes.io/docs/concepts/workloads/controllers/statefulset/#limitations

StatefulSets currently require a Headless Service to be responsible for the network identity of the Pods. You are responsible for creating this Service.

Add latency rate calculation task and related dashboard cells

[influxdb] documentation issue

Under the paragraph where it talks about starting with auth enabled, it states;

In values.yaml, change .Values.config.http.auth_enabled to true.

With the new chart, this should be auth-enabled instead of auth_enabled. auth_enabled will do nothing, as it's copied into the influxdb config 1 on 1.

Chart repository empty

Hello,

All charts from github pages repository have been removed:
eeb10e4

I didn't look how helm/chart-releaser-action works but It can be related to the github pages down:
https://www.githubstatus.com/

Thanks !

Monitor chronograf

Hello,

Is there any way we can set up monitor chronograf pods in kubernetes?
I tried to setup servicemonitor/podmonitor but I noticed that chronograf doesn't have the /metrics endpoint.
Could you please help me to setup a monitor on chronograf service/pod?
Thanks in advance

[influxdb-enterprise] issue with mounting a CA cert

Seeing an issue issue when trying to mount an internal CA cert. The cert itself seems to be working, but license validation is throwing error:

Reading license locally from: /tmp/influx-enterprise.key.json
2020/07/15 16:40:39 Error trying to fetch license: Get https://portal.influxdata.com/api/v1/licenses/269d13ff-4213-4c46-91c3-9bfa37a23886: x509: certificate signed by unknown authority

initScripts is incompatible with http-enabled: true

Figuring out how to both enable InfluxDB authentication and use an init script was a cumbersome process and I'm not happy with the solution I found.

Readme tells (https://github.com/influxdata/helm-charts/tree/master/charts/influxdb#starting-with-authentication):

To handle this setup on startup, a job can be enabled in values.yaml by setting .Values.setDefaultUser.enabled to true.

SetDefaultUser is an asynchronous process run in a dedicated container. Init scripts are run by container entrypoint. If admin user is created after init script is executed, commands are rejected because there is no admin user. If admin user is created before init script is executed, commands are rejected because init script doesn't provide, nor know, the right credentials.

It also not possible to create an admin user from the init script.

I figured out it was possible to set INFLUXDB_ADMIN_(USER|PASSWORD) environment variables. It this case, entrypoint creates the admin user first and then provides the proper credentials to subsequent commands.

This leads me to ask two questions:

• Shouldn't set-auth post-install hook be superseded by another mechanism ?
• Is there is way to use secrets to set INFLUXDB_ADMIN_PASSWORD ?

add support to backup data on s3 aws.

Currently, backups are only supported for azure and gcs. I want to add support for data backup on Amazon s3. Similar issue was created on helm repo.

Telegraf: Unable to pass multiple tags options to config.processors.regex

I am trying to define multiple regex rules to convert data, all operating on the tags. The telegraf docs show you can supply multiple tags or fields as subsequent blocks, and we have this running in an environment where we direct control of the telegraf.conf.

This config is being defined in values.yaml as follows:

processors:
    - regex:
        namepass: mqtt_consumer
        tags:
          key: "topic"
          pattern: '''^([A-Za-z0-9\-\.]+\/){0}([A-Za-z0-9\-\.]+)[A-Za-z0-9\-\.\/]*$'''
          replacement: "${2}"
          result_key: "domain"
        tags:
          key: "topic"
          pattern: '''^([A-Za-z0-9\-\.]+\/){1}([A-Za-z0-9\-\.]+)[A-Za-z0-9\-\.\/]*$'''
          replacement: "${2}"
          result_key: "location"
        ....
    - strings:
        trim_prefix:
          field_key: "*"
          prefix: "points_"
        trim_suffix:
          field_key: "*"
          prefix: "_present_value"

Expected output:

[[processors.regex]]
    namepass = ["mqtt_consumer"]

  [[processors.regex.tags]]
    key = "topic"
    pattern = '''^([A-Za-z0-9\-\.]+\/){0}([A-Za-z0-9\-\.]+)[A-Za-z0-9\-\.\/]*$'''
    replacement = "${2}"
    result_key = "domain"

  [[processors.regex.tags]]
    key = "topic"
    pattern = '''^([A-Za-z0-9\-\.]+\/){1}([A-Za-z0-9\-\.]+)[A-Za-z0-9\-\.\/]*$'''
    replacement = "${2}"
    result_key = "location"
.....

[[processors.strings]]

  [[processors.strings.trim_prefix]]
    field_key = "*"
    prefix = "points_"

  [[processors.strings.trim_suffix]]
    field_key = "*"
    suffix = "_present_value"

Actual output:

[[processors.regex]]
  namepass = "mqtt_consumer"
  [[processors.regex.tags]]
    key = "topic"
    pattern = "'^([A-Za-z0-9\\-\\.]+\\/){4}([A-Za-z0-9\\-\\.]+)[A-Za-z0-9\\-\\.\\/]*$'"
    replacement = "${2}"
    result_key = "udmi_type"
[[processors.strings]]
  [[processors.strings.trim_prefix]]
    field_key = "*"
    prefix = "points_"
  [[processors.strings.trim_suffix]]
    field_key = "*"
    prefix = "_present_value"

There is only one processors.regex.tags block defined, being the last in the values.yaml, and I believe in the Helm interpolation is overwriting the blocks with a matching tags key.

Is this a limitation/bug in the Helm charts, or with my yaml configuration in the values.yaml?

Telegraf helm Chart:- volumeMounts issue [with helm3 strict validation]

hi
ERROR:-
Error: unable to build kubernetes objects from release manifest: error validating "": error validating data: ValidationError(Deployment.spec.template.spec.volumes[1]): unknown field "mountPath" in io.k8s.api.core.v1.Volume

Getting the above error while trying to deploy with helm3.
I have added the volumeMounts section in telegraf values.yaml.

volumeMounts:

• name: tls-config
  mountPath: /app/certs/
  configMap:
  name: "telegraf-agent-configmap"

Where this is been called in 2 places (under volumes and volumeMounts) in deployment.yaml as shown below.

    volumeMounts:
    {{- range .Values.volumeMounts}}
    - name: {{ .name }}
      mountPath: {{ .mountPath }}
    {{- end }}

  volumes:
  {{- if .Values.volumeMounts }}

{{ toYaml .Values.volumeMounts | indent 6 }}
{{- end}}

But helm3 does a strict validation where in it cannot accept mountPath in volumes and in volumeMounts mountPath turns out to be a mandatory field.

Kindly confirm when this issue will be fixed.

[telegraf] No support for `[[inputs.postgresql_extensible.query]]`?

Hi everybody,

currenty I'm migrating my local telegraf docker instance to a helm instance.
I'm currently using the input plugin postgresql_extensible with which I can define some queries in this format:

[[inputs.postgresql_extensible]]
  address = "..."

  [[inputs.postgresql_extensible.query]]
    script="/app/foobar.sql"
    version="1200"
    withdbname=false
    tagvalue="tenant"
    measurement="something"

This is my current yaml version of it:

config:
  - inputs:
    - postgresql_extensible:
      address: "..."
      query:
        script: "/app/foobar.sql"
        version: "1200"
        withdbname: "false"
        tagvalue: "tenant"
        measurement: "something"

If I try to run this I'm getting this error message:

2020-06-18T12:10:31Z I! Starting Telegraf 1.14.4
2020-06-18T12:10:31Z I! Using config file: /etc/telegraf/telegraf.conf
2020-06-18T12:10:31Z E! [telegraf] Error running agent: Error parsing /etc/telegraf/telegraf.conf, line 75: cannot unmarshal TOML table into postgresql_extensible.query (need struct or map)

This is the produced TOML at that possition:

[[inputs.postgresql_extensible]]
  address = "..."
  [inputs.postgresql_extensible.query]
    measurement = "something"
    script = "/app/foobar.sql"
    tagvalue = "tenant"
    version = "1200"
    withdbname = "false"

As you can see there is just one bracket pair ([...query]) and not two ([[...query]]).
I've looked into the _helpers.tpl and for me it looks like, that this use case isn't support currently.

Am I correct and does someone know a workaround or something?

Make istio template video

Support Deployment of the InfluxDB & Telegraf Operators

Port 8088 is blocked

I am unable to upload a backup because the admin port is blocked.

[telegraf]: unable to pass `processor.converter.field` values

Telegraf Helm chart version: 1.7.7
Telegraf Docker image 1.13-alpine
Amazon Linux on EKS

We want to convert a field (system_location_sectionFloorNumber) into a tag. The following config works when passed as direct yaml as part of a telegraf.conf file:

[[processors.enum]]
  order = 1
  [[processors.enum.mapping]]
    dest = "system_location_sectionFloorNumber"
    field = "system_location_section"
    [processors.enum.mapping.value_mappings]
        1st = 1
        2nd = 2
        3rd = 3
        4th = 4
        5th = 5
        Basement = -1
        Ground = 0
[[processors.converter]]
  order: 2
  [[processors.converter.fields]]
    tag: [ "system_location_sectionFloorNumber" ]

When included as the following, as part of the values.yaml that is passed to the Helm chart of the Telegraf Docker image, it returns the following error:

  processors:
    - enum:
        order: 1
        mapping:
          field: system_location_section
          dest: system_location_sectionFloorNumber
          value_mappings:
            2nd: 2
            1st: 1
            Ground: 0
            Basement: -1
    - converter:
        order: 2
        fields:
          tag: system_location_sectionFloorNumber

E! [telegraf] Error running agent: Error parsing /etc/telegraf/telegraf.conf, line 39: (converter.Converter.Fields) cannot unmarshal TOML array table into converter.Conversion (need slice)

This is the case whether the tag key is passed as
tag: [ system_location_sectionFloorNumber ]

tag:
  - system_location_sectionFloorNumber

etc

Error installing on Kubernetes

I am getting the following error

Error: unable to build kubernetes objects from release manifest: error validating "": error validating data: ValidationError(Service.spec.ports[0].port): invalid type for io.k8s.api.core.v1.ServicePort.port: got "string", expected "integer"

telegraf-ds: allow adding priorityclass

Could it possible to add an optional priorityclass?
This will prevent that ds remains unscheduled if the node has not enough resources:
https://kubernetes.io/docs/concepts/configuration/pod-priority-preemption/

Thanks

Will there be major changes in helm chart for deploying Influx 2?

When will charts be upgraded to influx 2?

Does image upgrade to 2 work or will new chart be created?

Write Istio Template Instructions

Make ingress path configurable

Hi,

Would it be possible to make the influxdb chart Ingress path configurable from the values file? At the moment the value is hard coded into "/".

I have some use cases to change the path and also it does not work well with aws-alb-ingress-controller, which then creates an Application Load Balancer, that only forwards request from '/', so calling /query does not work.

`
{{- end }}
rules:

• host: {{ .Values.ingress.hostname }}
  http:
  paths:
  • path: /
    backend:
    serviceName: {{ include "influxdb.fullname" . }}
    servicePort: 8086
    {{- end -}}
    `

[influxdb-enterprise] HTTPS service call returning unknown protocol

When making a curl call from any node in the cluster of a data node using HTTPS to port:8086 , the service returns SSL23_GET_SERVER_HELLO:unknown protocol.

• TLS enabled

Fix Usage of "stable/*" on Individual Chart READMEs

[Enterprise] HTTPs Misconfigured for Data HTTP Port

When I wrote this chart, I thought that HTTPs only needed enabled on the [cluster] section of the TOML.

However, it seems that the HTTP API, port 8086, needs this configured separately.

This went unnoticed due to the port mismatched that was fixed yesterday: #136

https://docs.influxdata.com/enterprise_influxdb/v1.8/administration/config-data-nodes/#cluster-settings

These HTTPs settings need duplicated in the data-configmap under [http]

Using the same x509s as the cluster settings is OK for now and will reduce complexity of values.yaml, but it's not a wild option to support different DNS names for RPC and HTTP. ... eventually

telegraf-ds: support exposing secret key as env

This will make configuring InfluxDB 2 much easier

secretToEnv:
 - name: my-secret
   key: my-key
   env: TOKEN

The collectd service uses TCP when it should use UDP

The collectd service template does not specify the UDP protocol and therefore launches as TCP. However, the collectd service uses UDP and not TCP. Manually editing the deployment to use UDP allowed traffic to flow in.

telegraf-ds: allow configuration of kubernetes port

Currently hardcoded to 10250, would be great to be able to use a custom value.

Thanks!

Support Node Monitoring Through DaemonSet

This could be copied from https://github.com/influxdata/tick-charts/tree/master/telegraf-ds

Make Production Hud (ISTIO) Generic and Pluggable

make it so that you can add cells when adding new production environments

Chronograf Chart Probes Incompatible with BASE_PATH & Ingress other than /

Using the updated configurable ingress path in the helm chart is great in conjunction with the BASE_PATH which tells chronograf to serve from a sub-path, like /chronograf. Unfortunately the probes do not support this leaving the user with a CrashLoopBackOff due to failure of the probes.

       livenessProbe:
          httpGet:
            path: /ping
            port: api
        readinessProbe:
          httpGet:
            path: /ping
            port: api

Recommendation is to add a chart value that will set both the env: for BASE_PATH, and the /{BASE_PATH}/ping to the same base path, and everything will work great if you are using a proxy, or something like nginx-ingress.

telegraf chart: unable to specify input tags

I'm unable to specify tags on my inputs using the telegraf chart.

In my values.yml I have something like this:

  inputs:
  - prometheus:
      kubernetes_services: 
      - http://blackbox-exporter:9115/probe?....
      metric_version: 2
      tags:
        foo: bar

Which gets templated into the following TOML:

[[inputs.prometheus]]
  alias = "blackbox"
  kubernetes_services = [
    "http://blackbox-exporter:9115/probe?...
  ]
  metric_version = 2
  [[inputs.prometheus.tags]]
    foo = "bar"

However, tags is a table, not an array, so I think it should be [inputs.prometheus.tags] instead of [[inputs.prometheus.tags]]. Am I doing something wrong or is the template incorrect? Any workarounds?

helm-charts/charts/telegraf/templates/_helpers.tpl

Line 238 in 4eb7881

[[inputs.{{ $input }}.{{ $key }}.{{ $k }}]]

Make errorRateCalculation Task Generic

currently refers to "acc" environment