qarik-group / bootstrap-kubernetes-demos Goto Github PK

The deploy step gets stuck at the end:

11:27:00PM: ---- waiting on 1 changes [298/299 done] ----
11:27:06PM: ongoing: reconcile service/istio-ingressgateway (v1) namespace: istio-system
11:27:06PM: ^ Load balancer ingress is empty
11:28:01PM: ---- waiting on 1 changes [298/299 done] ----
11:28:06PM: ongoing: reconcile service/istio-ingressgateway (v1) namespace: istio-system
11:28:06PM: ^ Load balancer ingress is empty
11:29:02PM: ---- waiting on 1 changes [298/299 done] ----
11:29:07PM: ongoing: reconcile service/istio-ingressgateway (v1) namespace: istio-system
11:29:07PM: ^ Load balancer ingress is empty
11:30:02PM: ---- waiting on 1 changes [298/299 done] ----
11:30:08PM: ongoing: reconcile service/istio-ingressgateway (v1) namespace: istio-system
11:30:08PM: ^ Load balancer ingress is empty
11:31:03PM: ---- waiting on 1 changes [298/299 done] ----
11:31:08PM: ongoing: reconcile service/istio-ingressgateway (v1) namespace: istio-system
11:31:08PM: ^ Load balancer ingress is empty
11:32:03PM: ---- waiting on 1 changes [298/299 done] ----
11:32:08PM: ongoing: reconcile service/istio-ingressgateway (v1) namespace: istio-system
11:32:08PM: ^ Load balancer ingress is empty
11:33:04PM: ---- waiting on 1 changes [298/299 done] ----
11:33:09PM: ongoing: reconcile service/istio-ingressgateway (v1) namespace: istio-system
11:33:09PM: ^ Load balancer ingress is empty

Currently clean first runs down. This is unexpected.

Defer to bootstrap-infrastructure-xyz status or similar to determine if cluster running or not.

For SCF it can take a long time for cf-operator to create scf pods, and for the scf pods to start running healthily. bootstrap-system-scf watch could run watch kubectl -n scf get pods.

I just ran the script (Azure version) to play with Knative with the following command bootstrap-kubernetes-demos up --azure --knative and noticed one thing:

• The timeout call to validate knative setup kubectl wait -l app=controller pod --for condition=Ready -n knative-serving --timeout 30sis too short and the script will exit. It took almost 5 minutes until this pod was up and the condition met

If i run the boostrap-kubernetes-demos script under linux (in my case linux mint) i get weird outputs like this:

bootstrap-kubernetes-demos up --az basename: extra operand 'state/systems/scf' Try 'basename --help' for more information. basename: extra operand 'state/systems/scf' Try 'basename --help' for more information. basename: extra operand 'state/systems/scf' Try 'basename --help' for more information. basename: extra operand 'state/systems/scf' Try 'basename --help' for more information. basename: extra operand 'state/systems/scf' Try 'basename --help' for more information. basename: extra operand 'state/systems/scf' Try 'basename --help' for more information. basename: extra operand 'state/systems/scf' Try 'basename --help' for more information. basename: extra operand 'state/systems/scf' Try 'basename --help' for more information. basename: extra operand 'state/systems/scf' Try 'basename --help' for more information. basename: extra operand 'state/systems/scf' Try 'basename --help' for more information.

you can reproduce this within a docker ubuntu container:

$ docker run -it -v $PWD:/tmp/test ubuntu
root@735ba4641c26:/# cd /tmp/test/

root@735ba4641c26:/tmp/test# bootstrap-kubernetes-demos up --az
basename: extra operand 'state/systems/scf'
Try 'basename --help' for more information.
basename: extra operand 'state/systems/scf'
...... 

I run: bootstrap-kubernetes-demos up --azure --scf

CF-Operator is deployed correct:
`--> Using cf-operator-v0.4.2+85.gc6d71da5.tgz

• helm upgrade --install --wait --namespace scf cf-operator https://cf-operators.s3.amazonaws.com/helm-charts/cf-operator-v0.4.2%2B85.gc6d71da5.tgz --set provider=gke --set operator-webhook-use-service-reference=true
  Release "cf-operator" does not exist. Installing it now.
  NAME: cf-operator
  `

but then helm stucks on eirini:
`Install Cloud Foundry/Eirini (scf) for scf.suse.dev
--> Using scf-3.0.0-c656daa.tgz

• helm upgrade --install --namespace scf scf https://scf-v3.s3.amazonaws.com/scf-3.0.0-c656daa.tgz --set system_domain=scf.suse.dev --set features.eirini.enabled=true

^C
immi@immi-V`
i waited aprox. 15 min and then hitted crtl-c

Currently the state of which infrastructure/systems/configuration have been selected are stored within the project's state/ folder.

To allow users to have multiple deployments, or to allow them to commit their state/ folder to a git repo outside of this git repo, we should allow an alternate state folder to be provided by env var.

For example https://github.com/starkandwayne/bucc/blob/master/bin/bucc#L17

Just encountered this issue on an Azure K8s (v 1.15.7) environment

kubectl apply --wait=true --filename https://raw.githubusercontent.com/knative/serving/master/third_party/istio-1.3.6/istio-lean.yaml
error: unable to read URL "https://raw.githubusercontent.com/knative/serving/master/third_party/istio-1.3.6/istio-lean.yaml", server reported 404 Not Found, status code=404

Instead of failing if kwt missing, download it into bin/

hi,
i used the updated repo and get the following warning outputs :

bootstrap-kubernetes-demos up --google basename: zusätzlicher Operand »state/systems/scf“ „basename --help“ liefert weitere Informationen. basename: zusätzlicher Operand »state/systems/scf“ ... snip....

best practise for cf-operator is for it to go in its own namespace, and to watch kubecf namespace. Perhaps cf-operator-kubecf namespace -> kubecf namespace, to allow for multiple cf-operators -> watched namespace.

https://github.com/knative/serving/tree/master/third_party

When I bootstrap --scf there are sometimes additional admin steps to perform once scf is finally running. https://github.com/starkandwayne/update-all-cf-buildpacks is an example -- it blocks via an init container until SCF is running.

For example, I'd like to install service brokers once SCF is running.

hi, I am running into problems with the directions at the point of using ytt.

I reverted to using Dockerhub instead of GCloud repo, to get past the problems when parsing app-registry.yaml.

However now I am blocked on setting up istio_static_ip. When I had that value in static-ip.yaml, I was getting this:

ytt: Error: Overlaying data values (in following order: values/00-values.yml, values/10-images.yml, values/20-secrets-config-values.yml, values/30-experimental-values.yml, cf-values.yml, static-ip.yml):
Document on line static-ip.yml:2:
Map item (key 'istio_static_ip') on line static-ip.yml:3:
Expected number of matched nodes to be 1, but was 0

Then I tried just adding istio_static_ip directly to cf-values.yml at the end, but now I am getting:

ytt: Error: Overlaying data values (in following order: values/00-values.yml, values/10-images.yml, values/20-secrets-config-values.yml, values/30-experimental-values.yml, cf-values.yml):
Document on line cf-values.yml:2:
Map item (key 'istio_static_ip') on line cf-values.yml:345:
Expected number of matched nodes to be 1, but was 0

any ideas helpful