dsohk / susecap-tf-eks

Deploy SUSE CAP onto AWS EKS
MIT License
3 stars 7 forks source link

Cloudflare? #2

Open kevinayres opened 4 years ago

kevinayres commented 4 years ago

Derek, why use Cloudflare? For EKS testing, it's easiest to simply use R53. Please advise. Thanks. Kevin

dsohk commented 4 years ago

Kevin,

Short answer: cost consideration.

AWS Route53 requires “maintenance fees” on any domain hosted there and it charges the traffic going through it as well. Cloudflare is completely free to start with. Since one of my goals is to minimize cost of hosting SUSE CAP on AWS, any programmable with free-tier DNS services such as cloudflare is chosen.

Thanks, Derek

kevinayres commented 4 years ago

Thanks Derek! Makes sense

~Kevin

On Aug 26, 2019, at 5:53 PM, Derek So notifications@github.com<mailto:notifications@github.com> wrote:

Kevin,

Short answer: cost consideration.

AWS Route53 requires “maintenance fees” on any domain hosted there and it charges the traffic going through it as well. Cloudflare is completely free to start with. Since one of my goals is to minimize cost of hosting SUSE CAP on AWS, any programmable with free-tier DNS services such as cloudflare is chosen.

Thanks, Derek

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/dsohk/susecap-tf-eks/issues/2?email_source=notifications&email_token=ACLFQZPE2UEPQ5T4BRWK27LQGR3FBA5CNFSM4IPBXLKKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD5GDVZY#issuecomment-525089511, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ACLFQZMXHQ7QOBEWBAIA7XLQGR3FBANCNFSM4IPBXLKA.

kevinayres commented 4 years ago

Derek, I get the following error when executing script 03:

./03-install-suse-cap.sh: line 93: jq: command not found An error occurred (ResourceNotFoundException) when calling the DescribeCluster operation: No cluster found for name: susecap-eks.

Interestingly, the cluster creates successfully but describe-cluster fails and also “aws eks list-clusters” returns null. I’m authenticating correctly (i.e. “aws s3 ls” works.) Have you experienced this with aws eks command?

The pods don’t appear to be there:

bash-3.2$ kubectl get pods

No resources found.

The cluster is alive:

bash-3.2$ kubectl cluster-info

Kubernetes master is running at https://104.198.11.124

GLBCDefaultBackend is running at https://104.198.11.124/api/v1/namespaces/kube-system/services/default-http-backend:http/proxy

Heapster is running at https://104.198.11.124/api/v1/namespaces/kube-system/services/heapster/proxy

KubeDNS is running at https://104.198.11.124/api/v1/namespaces/kube-system/services/kube-dns:dns/proxy

Metrics-server is running at https://104.198.11.124/api/v1/namespaces/kube-system/services/https:metrics-server:/proxy

~ Kevin

From: Derek So notifications@github.com Reply-To: dsohk/susecap-tf-eks reply@reply.github.com Date: Monday, August 26, 2019 at 5:53 PM To: dsohk/susecap-tf-eks susecap-tf-eks@noreply.github.com Cc: Author author@noreply.github.com, Kevin Ayres kevin.ayres@suse.com Subject: Re: [dsohk/susecap-tf-eks] Cloudflare? (#2)

Kevin,

Short answer: cost consideration.

AWS Route53 requires “maintenance fees” on any domain hosted there and it charges the traffic going through it as well. Cloudflare is completely free to start with. Since one of my goals is to minimize cost of hosting SUSE CAP on AWS, any programmable with free-tier DNS services such as cloudflare is chosen.

Thanks, Derek

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/dsohk/susecap-tf-eks/issues/2?email_source=notifications&email_token=ACLFQZPE2UEPQ5T4BRWK27LQGR3FBA5CNFSM4IPBXLKKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD5GDVZY#issuecomment-525089511, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ACLFQZMXHQ7QOBEWBAIA7XLQGR3FBANCNFSM4IPBXLKA.

zoopster commented 4 years ago

jq --> https://stedolan.github.io/jq/ jq is used to extract the cluster endpoint. You can simply run 'aws describe cluster --name ' and get the end point.

kubectl get pods will default to the "default namespace" which has no running pods by default use 'kubectl get pods -n kube-system'