netrisai / netris-operator Goto Github PK

View Code? Open in Web Editor NEW

40.0 5.0 3.0 720 KB

The Kubernetes Operator for Netris

License: Apache License 2.0

Dockerfile 0.22% Makefile 0.89% Go 97.73% Smarty 0.69% Python 0.46%

kubernetes kubernetes-operator netops infrastructure infrastructure-as-code netris networking

netris-operator's Introduction

Netris Operator

Netris Operator is a Kubernetes Operator for managing and automating tasks related to managing Netris VPC.

Netris automatically operates the physical switching network in your data center, bare metal cloud, or at the edge. Netris automatically enables routing, elastic load balancer, firewall, DHCP, NAT, and other essential network services utilizing ordinary servers and SmartNICs. Netris makes your network automatic, resilient, and uncomplicated - just like the cloud but very cost-efficient.

Deploying the Operator

The current version of netris-operator requires Kubernetes >= 1.16.

GitOps

Kubernetes manifests are located in the deploy/ folder. To deploy the operator manually using Kubernetes manifests or to integrate it into your GitOps flow please follow these instructions.

Helm Charts

You can also use Helm charts to deploy Netris Operator.

Documentation

Netris Operator managing resources samples are available in the samples/ folder.

Compatibility with Netris-Controller

Operator version	Controller version
`v0.X`	`v2.9`
`v1.X`	`v3.0`
`v2.X`	`v3.1+`
`v3.X`	`v4.1+`

Features

Managing Netris Controller via CRD
Automatically creating L4LB resource for type: load-balancer services
All CNIs are welcome

netris-operator's People

Contributors

Stargazers

Watchers

Forkers

ekmixon saswatp ixodie

netris-operator's Issues

Calico integration

Services status remain provisioning

I installed Netris Operator in Kubernetes cluster and Istio ingress. Services show up in Netris controller UI in services ->L4 load balancer but all remain in "provisioning" status.

stale state in operator after UI delete

Hi, I've create a vnet via k8s CR, and deleted it in web UI. According to the docs, I was expecting for it to be re-created in the controller DB (and visible in web UI). Instead I see it in k8s as:

air-1   active              Default   Admin   Failure        8m7s

and the following is in the logs

{"level":"info","ts":1643119788.7306645,"logger":"VNet","msg":"Go to delete","name":"netris/air-1"}
{"level":"error","ts":1643119788.738838,"logger":"VNet","msg":"","name":"netris/air-1","error":"{deleteVNet} {deleteVNet} Please provide a valid v-net."}
{"level":"error","ts":1643119788.768768,"logger":"VNetMeta","msg":"","name":"netris/air-1","error":"{GetVNetsByID} {GetByID} {http.Get} {\"data\":null,\"errors\":{\"error\":\"Please provide a valid v-net.\"},\"message\":\"Please provide a valid v-net.\",\"isSuccess\":false,\"meta\":{\"apiVersion\":\"v2\",\"statusCode\":500}}"}
{"level":"info","ts":1643119788.7732,"logger":"VNet","msg":"Go to delete","name":"netris/air-1"}
{"level":"error","ts":1643119788.7800682,"logger":"VNet","msg":"","name":"netris/air-1","error":"{deleteVNet} {deleteVNet} Please provide a valid v-net."}
{"level":"info","ts":1643119788.783426,"logger":"VNet","msg":"Go to delete","name":"netris/air-1"}
{"level":"error","ts":1643119788.7898724,"logger":"VNet","msg":"","name":"netris/air-1","error":"{deleteVNet} {deleteVNet} Please provide a valid v-net."}

ImagePullBackOff tag HEAD not found

after applying netris-operator.yaml, pod didn't run - ImagePullBackOff error
Failed to pull image "netrisai/netris-operator:HEAD"

panic: runtime error: index out of range [0] with length 0

I'm installing netris-operator on a new cluster (without any calico resources) and seeing operator in CrashLoopBackOff with the following log:

{"level":"error","ts":1643045695.4067953,"logger":"LBWatcher","msg":"","error":"CRD Not found"}
I0124 17:34:55.406862       1 leaderelection.go:242] attempting to acquire leader lease  netris/abac3abe.netris.ai...
{"level":"info","ts":1643045695.4070582,"logger":"controller-runtime.manager","msg":"starting metrics server","path":"/metrics"}
panic: runtime error: index out of range [0] with length 0

goroutine 196 [running]:
github.com/netrisai/netris-operator/calicowatcher.(*Watcher).deleteProcess(0xc00029eb40, 0x16823fb, 0x15)
        /workspace/calicowatcher/main.go:319 +0xb29
github.com/netrisai/netris-operator/calicowatcher.(*Watcher).mainProcessing(0xc00029eb40, 0xc00004007b, 0x15)
        /workspace/calicowatcher/main.go:384 +0x1fd
github.com/netrisai/netris-operator/calicowatcher.(*Watcher).start(0xc00029eb40)
        /workspace/calicowatcher/main.go:137 +0x128
github.com/netrisai/netris-operator/calicowatcher.(*Watcher).Start(0xc00029eb40)
        /workspace/calicowatcher/main.go:159 +0x1e8
created by main.main
        /workspace/main.go:187 +0x1205

Steps to reproduce (assuming netris is already installed):

kubectl -n netris create secret generic netris-creds \
  --from-literal=host="http://10.110.50.235" \
  --from-literal=login="netris" --from-literal=password="newNet0ps"

helm install netris-operator netrisai/netris-operator \
--namespace netris