https://metaplay.io logo
Powered by
Metrics Dashboard not available
# questions-and-feedback
a
Hi, should the metrics dashboard be available by default? I see the 'Game server metrics' link, and have the game-admin role (view everything in dashboard), but then Grafana gives me: Dashboard not found Error You'll need additional permissions to perform this action. Permissions needed: dashboards:read The logs view works ok.
v
Hi! The dashboard being missing is a bug and it has been fixed in metaplay-gameserver Helm chart v0.6.0
The quickest fix would be making a new game server deployment with metaplay-gameserver Helm chart v0.6.0 and you should be able to see the dashboard in Grafana. A command snippet for your reference:
Copy code
helm upgrade --install --repo https://charts.metaplay.dev --version 0.6.0 -n imaginedesign-<environment name> -f deployment.yaml gameserver metaplay-gameserver
a
i failed with that so far, what's deployment.yaml ? i don't have it. i have a kubeconfig.yaml from an old manual deploy but that didn't work with it, it tries to connect to localhost .. maybe i need to have some server running locally? i don't know basically anything about helm nor kubernetes
maybe it needs the docker deamon, that seems to be have some problem here now, i'll restart..
v
The 
deployment.yaml
file is a yaml manifest that will be read by Helm for populating customized values which will be later consumed by the metaplay-gameserver Helm chart. May I know which tenant environment are you experiencing the dashoard missing issue? Is it 
imaginedesign-develop
or 
imaginedesign-stable
?
a
i failed locally but i saw that the helm-chart-version was also in the workflow we use for CI on GitHub so i upgraded it there and am betting will have better luck nw
this was in develop, but i suppose it's the same in both
v
Yep, probably GitHub Action is the place you want to check with game server deployment, and upgrading the chart versions there should solve the problem, you are likely on the right track 🙂
a
yah. i did deploy manually initially but have been happy with the GH action since .. i guess i just missed some step now, did try to run the auth and kubernetes setups etc but finally got stuck at Error: Kubernetes cluster unreachable: Get "https://6CD52F69F68708A186C714E22FCCBBF8.gr7.eu-west-1.eks.amazonaws.com/version": getting credentials: exec: executable metaplay-auth failed with exit code 1
v
The above error sounds like an authentication failure with the metaplay-auth CLI tool. Did this error pop up in the CI or locally?
The 
Kubernetes cluster unreachable
error is due to missing a valid 
kubeconfig.yaml
file which is essential for communicating with the EKS cluster. In your local machine, you can obtain the 
kubeconfig.yaml
with the metaply-auth CLI tool, here's how:
Copy code
bash
export KUBECONFIG=/path/to/<my-kubeconfig-file-location>
metaplay-auth get-kubeconfig metaplay-imaginedesign-develop --output $KUBECONFIG
a
that was locally. i did run that kind of command first but perhaps made some mistake
so in CI it worked and now i see the dashboard, great, thanks!
hm now the LiveOps Dashboard gives error 500 though 😮
i try reverting the change to get it back for now
v
Hi! We have noticed some errors in the 
external-auth-server
which is a module that controls the authentication to both LiveOps Dashboard and Grafana. The 
external-auth-server
has been recreated in 
imaginedesign-develop
and now you should be able to access the Dashboard.
The metaplay-gameserver v0.6.0 Helm chart may still have some bugs and we have been trying to fix them ASAP. We are sorry for the inconvenience, and thank you for notifying us the issue!
We have just published metaplay-gameserver Helm chart v0.6.1 with a hotfix.🛠️ You can now use the latest chart version in your CI to make a new game server deployment, which could help you keep away from the 500 error.
PreviousNext
Powered by