Forward your temperature statistics to a time series database so you can look at real-time graphs of temperature over time.
Backup v1 influxdb:
kubectl exec -it influxdb -- bash
> cd /tmp/
> influxd backup -db nest_temperature_forwarder -portable backup-v1
> tar -cvf backup-v1.tar.gz backup-v1/
kubectl cp influxdb:/tmp/backup-v1.tar.gz ./backup-v1.tar.gz
You can use the deploy/migrate.sh
script to migrate a influxdb v1 to a v2.
When the browser opens rename the bucket without autogen
.
A new backup named backup-v2.tar.gz
will be available to restore a clean helm
install from.
kubectl cp ./backup-v2.tar.gz influxdb:/tmp/backup-v2.tar.gz
kubectl exec -it influxdb -- bash
> export INFLUX_TOKEN=<TOKEN>
> export INFLUX_ORG=nest_temperature_forwarder
> cd /tmp/
> tar -xvf backup-v2.tar.gz backup-v2/
> influx restore --full backup-v2/
Older versions of nest_temperature_forwarder
ran as a kubernetes cronjob.
It now runs as a deployment with a scheduler in python code. This is to save
pod churn on smaller kubernetes clusters.
When running as a cronjob make sure to add --once
.
Usually nest_temperature_forwarder
should be run in a kubernetes cluster.
However it can be run in any environment using environment variables and flags
to configure it.
usage: python temperature_forwarder.py [-h] [--health-check] [--health-check-path HEALTH_CHECK_PATH] [--health-check-delta HEALTH_CHECK_DELTA] [--delay-seconds DELAY_SECONDS] [--once] [--verbose]
Get metrics from the nest API and put them into influxdb.
optional arguments:
-h, --help show this help message and exit
--health-check Do get metrics by check last data point witin range (default: False)
--health-check-path HEALTH_CHECK_PATH
Path on disk to store last successful run time (default: /tmp/healh_check.txt)
--health-check-delta HEALTH_CHECK_DELTA
Number of minutes behind before failing healthcheck (default: 20)
--delay-seconds DELAY_SECONDS
Seconds between data points (default: 300)
--once Do not start scheduler. Get and store single data point (default: False)
--verbose Enable verbose output (default: False)
Name | Description | Default |
---|---|---|
POSTAL_CODE |
A UK postal code to always get weather data for. Uses nest thermostat postal code by default. | "" |
NEST_ACCESS_TOKEN |
API token for nest. If blank nest api not queried. | "" |
OPENWEATHERMAP_API_KEY |
API Key for https://openweathermap.org/. If blank not external temperature data is logged. | "" |
INFLUX_TOKEN |
API token for influxdb. If logged, output is only to stdout. | "" |
INFLUX_URL |
influxdb full url | http://localhost:8086 |
INFLUX_BUCKET |
influxdb bucket name | nest_temperature_forwarder |
INFLUX_ORG |
influxdb organization name | nest_temperature_forwarder |
Github Actions are triggered to build and deploy the app on release. Release Drafter is used to draft releases based on pull request titles. A service account is used to deploy via helm3. This must be created first.
---
apiVersion: v1
kind: Secret
metadata:
name: nest-temperature-forwarder
namespace: nest
stringData:
nest_access_token: ""
openweathermap_api_key: ""
# Admin user used for both influx and grafana
admin-user: "admin"
admin-password: ""
admin-token: ""
---
apiVersion: v1
kind: Secret
metadata:
name: grafana-datasources
namespace: nest
labels:
grafana_datasource: "true"
stringData:
datasources.yaml: |
apiVersion: 1
datasources:
- name: influxdb
type: influxdb
access: proxy
url: http://nest-influxdb2:80
secureJsonData:
token: "MATCH TOKEN ABOVE"
jsonData:
version: Flux
organization: nest_temperature_forwarder
defaultBucket: nest_temperature_forwarder
To create the service account and permissions, a cluster-admin needs to apply the following:
kubectl apply -f deploy/serviceaccount.yaml
Secrets for github actions are as follows:
DOCKER_TOKEN
: The PAT token for ghcr.ioK8S_SECRET
: The full yaml secret for the serviceaccountK8S_URL
: The url of the kubernetes api server