graphistry-helm developer docs

This document is for Graphistry helm contributors on tasks like setting up a local k8s, running Graphistry locally, submitting patches, and making new releases

For using the Graphistry helm charts in your own Kubernetes, See the main README

For an even more encompassing list of commands and options, see the kubectl cheatsheet

TO-DOs

~~calico rules~~
env secrets
autoscaling
tls intercontainer termination
some sort of observability down the line with Prometheus and grafana with traefik when we begin to phase out nginx in favor of traefik
k8s api dashboard
add logging
add multinode volume support with NFS (longhorn)

Convenient Commands

get pods

kubectl get pods

describe a pod

kubectl describe pod <name>

delete a helm chart

helm delete <chart name>

install a helm chart

helm install <chart name> ./<chart dir>

upgrade a chart

helm upgrade <chart name> ./<chart dir>

get a bash shell on a pod

kubectl exec -i -t <name> -- /bin/bash

get svc by name

kubectl get service nginx

port forward a pod

kubectl port-forward deployment/nginx 8000:80

get last 100 logs of a pod

kubectl logs --tail=100 deployment/nginx

get logs of a pod and follow

kubectl logs --follow deployment/nginx

add gpu daemonset to cluster

kubectl create -f https://raw.githubusercontent.com/NVIDIA/k8s-device-plugin/master/nvidia-device-plugin.yml

once daemonset has been installed and started if successful will see nvidia.com/gpu in nodes capacity here
kubectl get nodes -ojson | jq .items[].status.capacity

upgrade with a test run to check that charts are working

helm upgrade --dry-run graphchart-release ./<chart dir>

lint helm charts

helm lint <chart dir>

print env in a pod

kubectl exec envar-demo -- printenv

how to restart a pod

kubectl scale deployment nginx --replicas=0 && kubectl scale deployment nginx --replicas=1

cat a file in a pod

kubectl exec -it deployments/nginx -- cat /etc/resolv.conf

convert conf to config map

kubectl create configmap nginx-conf --from-file=./default.conf

describe configmap config in yaml

kubectl get configmaps nginx-conf -o yaml

describe prettified configmap config

kubectl describe configmaps nginx-config | nl

kubectl get service -n kube-system

checks cluster dns ip

get kube secrets

kubectl get secrets

get namespace

kubectl get namespace

parse env values

yq e ".env[1]" ./charts/graphistry-helm/values.yaml

update yaml in place with yq

yq e -i '.env[1] = "cool"' ./charts/graphistry-helm/values.yaml

check all api resources can interact with

kubectl api-resources

get configmaps from specific namespace

kubectl -n kube-public get configmaps

get pods on all namespaces

kubectl get pods -A

get pods on a particular node

kubectl describe node <node-name>

check storage of pod

kubectl -n <namespace> exec <pod-name> df

delete pvcs

kubectl delete pvc --all

check storage on pod , includes pvc if mounted

kubectl -n <namespace> exec <pod-name> -- df -h

continuously run kubectl get pods

watch kubectl get pods

to get initcontainer logs

kubectl logs deployments/nginx -c nginx-init-streamgl-viz

get contexts for kubectl

kubectl config get-contexts

change kubectl context

kubectl config use-context <yourClusterName>

get nodes

kubectl get nodes

describe the specified node