problem in try to install/try in local host

[TarepSH]

Hope you are doing will..

I really spent a lot of time found the solution's for this issues, but I can't, so is time to ask for your help,
when I try to install Magda in local host using minikube but i get this message:
kubectl apply -f role-binding.yaml error: resource mapping not found for name: "srv-acc-admin" namespace: "" from "role-binding.yaml": no matches for kind "ClusterRoleBinding" in version "rbac.authorization.k8s.io/v1beta1" ensure CRDs are installed first

Could help on this.

ubuntu lts 20.04
Docker version 20.10.18

Best
Tarek

[t83714]

Hi @TarepSH,
What's the "kubectl apply -f role-binding.yaml" command for?
If you try to follow any install guide, could you please share the reference?
If you just want to try the latest version, you can try with the following commands:

# Add Magda Helm Chart Repo:
helm repo add magda-io https://charts.magda.io

# create a namespace "magda" in your cluster
kubectl create namespace magda

# install Magda version v2.0.1 to namespace "magda", turn off openfass function and expose the service via loadBalancer
helm upgrade --namespace magda --install --version 2.0.1 --timeout 9999s --set global.openfaas.enabled=false,magda-core.gateway.service.type=LoadBalancer magda magda-io/magda

You can find more info in this section:

https://github.com/magda-io/magda#to-try-the-last-version-with-prebuilt-images

[TarepSH]

Thanks @t83714 for your replay

Yes, that's exactly what I did.

helm upgrade --namespace magda --install --version 2.0.1 --timeout 9999s --set global.openfaas.enabled=false,magda-core.gateway.service.type=LoadBalancer magda magda-io/magda
Release "magda" does not exist. Installing it now.
Error: execution error at (magda/charts/magda-function-history-report/templates/function.yaml:3:4): `magda-function-history-report` requires `magda-core`.`openfaas` to deploy. Please turn on openfaas by setting `global.openfaas.enabled` to `true` or turn off `magda-function-history-report` to avoid this error.

So, I change global.openfaas.enabled=false to true. and I get:

`helm upgrade --namespace magda --install --version 2.0.1 --timeout 9999s --set global.openfaas.enabled=true,magda-core.gateway.service.type=LoadBalancer magda magda-io/magda
Release "magda" does not exist. Installing it now.
Error: unable to build kubernetes objects from release manifest: [resource mapping not found for name: "indexer-reindex" namespace: "" from "": no matches for kind "CronJob" in version "batch/v1beta1"
ensure CRDs are installed first, resource mapping not found for name: "registry-db-auto-vacuum-job" namespace: "" from "": no matches for kind "CronJob" in version "batch/v1beta1"
ensure CRDs are installed first]`

[t83714]

minikube by default will install the latest k8s but we have a few resources that still use old APIs and not compatible with k8s v1.25+
You probably can try version 1.22.15 by setting minikube config:

minikube config set kubernetes-version 1.22.15

By the way, Magda requires min. 7GB ram to run. You probably also want to adjust the default memory size for minikube:

minikube config set memory 7851 

Unfortunately, you still need to delete the existing minikube VM to apply the changes:

minikube delete
minikube start

Then the following command should work

helm upgrade --namespace magda --install --version 2.0.1 --timeout 9999s --set global.openfaas.enabled=true,magda-core.gateway.service.type=LoadBalancer magda magda-io/magda

[chrowe]

I tried this and after it runnings for a while using a bunch of CPU I get...

❯ helm upgrade --namespace magda --install --version 2.0.1 --timeout 9999s --set global.openfaas.enabled=true,magda-core.gateway.service.type=LoadBalancer magda magda-io/magda
Release "magda" does not exist. Installing it now.
W1215 15:23:12.960817   33031 warnings.go:70] batch/v1beta1 CronJob is deprecated in v1.21+, unavailable in v1.25+; use batch/v1 CronJob
W1215 15:23:12.964285   33031 warnings.go:70] batch/v1beta1 CronJob is deprecated in v1.21+, unavailable in v1.25+; use batch/v1 CronJob
W1215 15:23:13.529534   33031 warnings.go:70] batch/v1beta1 CronJob is deprecated in v1.21+, unavailable in v1.25+; use batch/v1 CronJob
W1215 15:23:13.533202   33031 warnings.go:70] batch/v1beta1 CronJob is deprecated in v1.21+, unavailable in v1.25+; use batch/v1 CronJob

NAME: magda
LAST DEPLOYED: Thu Dec 15 15:23:11 2022
NAMESPACE: magda
STATUS: deployed
REVISION: 1
TEST SUITE: None
NOTES:
Thank you for installing magda.

Your release is named magda.

To learn more about the release, try:

  $ helm status magda
  $ helm get all magda
❯
❯ helm status magda
Error: release: not found
❯ helm get all magda
Error: release: not found

[t83714]

Hi @chrowe
As we deployed the magda into namespace magda (with --namespace switch), we need to specify the namespace when run helm command. e.g.

helm --namespace magda status magda

[chrowe]

Ok, that worked. Thanks!

Still having issues getting connected though...
Looking at https://github.com/magda-io/magda#to-try-the-last-version-with-prebuilt-images

I see the next step is...
echo $(kubectl get svc --namespace magda gateway --template "{{ range (index .status.loadBalancer.ingress 0) }}{{ . }}{{ end }}")

But I get

error: error executing template "{{ range (index .status.loadBalancer.ingress 0) }}{{ . }}{{ end }}": template: output:1:10: executing "output" at <index .status.loadBalancer.ingress 0>: error calling index: index of untyped nil
Error executing template: template: output:1:10: executing "output" at <index .status.loadBalancer.ingress 0>: error calling index: index of untyped nil. Printing more information for debugging the template: template was: {{ range (index .status.loadBalancer.ingress 0) }}{{ . }}{{ end }} raw data was: {"apiVersion":"v1","kind":"Service","metadata":{"annotations":{"meta.helm.sh/release-name":"magda","meta.helm.sh/release-namespace":"magda"},"creationTimestamp":"2022-12-17T03:37:59Z","labels":{"app.kubernetes.io/managed-by":"Helm"},"managedFields":[{"apiVersion":"v1","fieldsType":"FieldsV1","fieldsV1":{"f:metadata":{"f:annotations":{".":{},"f:meta.helm.sh/release-name":{},"f:meta.helm.sh/release-namespace":{}},"f:labels":{".":{},"f:app.kubernetes.io/managed-by":{}}},"f:spec":{"f:allocateLoadBalancerNodePorts":{},"f:externalTrafficPolicy":{},"f:internalTrafficPolicy":{},"f:ports":{".":{},"k:{\"port\":80,\"protocol\":\"TCP\"}":{".":{},"f:name":{},"f:port":{},"f:protocol":{},"f:targetPort":{}}},"f:selector":{},"f:sessionAffinity":{},"f:type":{}}},"manager":"helm","operation":"Update","time":"2022-12-17T03:37:59Z"}],"name":"gateway","namespace":"magda","resourceVersion":"523","uid":"5828c4f5-9893-4e28-aaa0-53b06ba5098f"},"spec":{"allocateLoadBalancerNodePorts":true,"clusterIP":"10.102.90.13","clusterIPs":["10.102.90.13"],"externalTrafficPolicy":"Cluster","internalTrafficPolicy":"Cluster","ipFamilies":["IPv4"],"ipFamilyPolicy":"SingleStack","ports":[{"name":"http","nodePort":30625,"port":80,"protocol":"TCP","targetPort":80}],"selector":{"service":"gateway"},"sessionAffinity":"None","type":"LoadBalancer"},"status":{"loadBalancer":{}}} object given to template engine was: map[apiVersion:v1 kind:Service metadata:map[annotations:map[meta.helm.sh/release-name:magda meta.helm.sh/release-namespace:magda] creationTimestamp:2022-12-17T03:37:59Z labels:map[app.kubernetes.io/managed-by:Helm] managedFields:[map[apiVersion:v1 fieldsType:FieldsV1 fieldsV1:map[f:metadata:map[f:annotations:map[.:map[] f:meta.helm.sh/release-name:map[] f:meta.helm.sh/release-namespace:map[]] f:labels:map[.:map[] f:app.kubernetes.io/managed-by:map[]]] f:spec:map[f:allocateLoadBalancerNodePorts:map[] f:externalTrafficPolicy:map[] f:internalTrafficPolicy:map[] f:ports:map[.:map[] k:{"port":80,"protocol":"TCP"}:map[.:map[] f:name:map[] f:port:map[] f:protocol:map[] f:targetPort:map[]]] f:selector:map[] f:sessionAffinity:map[] f:type:map[]]] manager:helm operation:Update time:2022-12-17T03:37:59Z]] name:gateway namespace:magda resourceVersion:523 uid:5828c4f5-9893-4e28-aaa0-53b06ba5098f] spec:map[allocateLoadBalancerNodePorts:true clusterIP:10.102.90.13 clusterIPs:[10.102.90.13] externalTrafficPolicy:Cluster internalTrafficPolicy:Cluster ipFamilies:[IPv4] ipFamilyPolicy:SingleStack ports:[map[name:http nodePort:30625 port:80 protocol:TCP targetPort:80]] selector:map[service:gateway] sessionAffinity:None type:LoadBalancer] status:map[loadBalancer:map[]]]

I am on an M1 Mac and new to Helm and Kubernetes/minikube

[t83714]

1> First time deploy must take some time, please watch pod staus to ensure all initial jobs are done and all services are online

2> the command line is for grabbing the external IP of the gateway service.
The external IP might take some time before available.
Before the external IP is available, you will keep getting the same error.
You can also use kubectl -n magda get svc gateway to get service info in human-readable format and watch the external IP column.
Please note: on newer minikube platform, you need to run the tunnel to get external IP assigned.
Please see doc here: https://minikube.sigs.k8s.io/docs/handbook/accessing/#loadbalancer-access

3> You can also follow this doc to expose the Magda service in your local cluster via ingress with HTTPS access:
https://github.com/magda-io/magda/blob/master/docs/docs/how-to-setup-https-to-local-cluster.md

[chrowe]

Finally got this working.
For me, I think the extra step I needed to do was run minikube tunnel