https://rancher.com/ logo
Join Slack
Powered by
This message was deleted.
# hobbyfarm
a
This message was deleted.
m
• HobbyFarm is deployed. • I updated the Gargantua and Gargantua-Shell deployments to use the 
ebauman/gargantua:pr-154-4
image. • The DigitalOcean provisioner is deployed and in the same namespace as HF. • DO secret exists, is in the proper namespace, and is properly referenced in the environment. • The DO token has been tested and is known to work. • VM template exists and is configured for use with DigitalOcean. • Environment is configured, based on the configuration settings. • Environment has the correct VM template. • Environment has the label for the hobbyfarm DO provisioner, as well as environment specifics and template mappings as shown in the provisioner documentation. • Scenario is configured to use the DO VM set. • Courses are configured to use the DO VM set and scenarios.
I’m getting the following error when attempting to launch the scenario as a user:
w
Hey Ed - first off, thanks for trying out HF !
Can you look in the JS console in your browser and see if you can identify the return error from the API call? We need to surface that information a bit better but this is a quick way to get that information.
m
Sure, one sec
Failed to load resource: the server responded with a status of 404 (Not Found)
It’s going to my hf link: http://hf.192-168-68-201.sslip.io/vmclaim/dynamic-5dsymclila
w
Interesting. If you look in Kubernetes does that vmclaim exist?
m
Looking
Copy code
➜ k get virtualmachineclaims -A
No resources found
w
okay. This tells us something went wrong with creating your session, as you should have a vmclaim associated with it. I would initially try to log out of HF, log back in, and try again. If that fails, look at the API call to create a session and see what the return information is from that
(again, sorry these troubleshooting steps suck)
m
Ha, no worries. I’ll log out and back in. What’s the best way to look at the API call?
w
JS console again. I would also try using 
kubectl logs -f
on the gargantua pod to see if there are errors there
m
Copy code
E1002 15:40:03.965518       1 vmclaimcontroller.go:377] no suitable environment for Eddard (vmt-ypwmm5z3lq): no suitable environment found. capacity reached

E1002 15:40:03.965527       1 vmclaimcontroller.go:264] error processing vmc dynamic-nzblzan4sj, taint session: no suitable environment found. capacity reached

I1002 15:40:03.973974       1 vmclaimcontroller.go:253] vmclaim dynamic-nzblzan4sj is tainted.. cleaning it up

E1002 15:40:06.000122       1 vmclaimserver.go:101] did not find the right virtual machine claim ID
w
Can you share your environment manifest here
m
Copy code
apiVersion: <http://hobbyfarm.io/v1|hobbyfarm.io/v1>
kind: Environment
metadata:
  creationTimestamp: "2023-09-26T17:14:15Z"
  generation: 7
  labels:
    <http://hobbyfarm.io/provisioner|hobbyfarm.io/provisioner>: digitalocean
  name: env-kyqhisvpba
  namespace: hobbyfarm-system
  resourceVersion: "667813"
  uid: bcfcf3de-9b2f-44f6-900a-1111829539b1
spec:
  count_capacity:
    vmt-ypwmm5z3lq: 1
  display_name: Test
  dnssuffix: ""
  environment_specifics:
    region: nyc1
    token-secret: do-secret
  ip_translation_map: {}
  provider: digitalocean
  template_mapping:
    vmt-ypwmm5z3lq:
      image: ubuntu-20-04-x64
      size: s-1vcpu-512mb-10gb
  ws_endpoint: <http://ws.192-168-68-200.sslip.io|ws.192-168-68-200.sslip.io>
w
Does your scenario require more VMs than are specified in count_capacity?
m
Let me double check
From what I can tell, I’m only requesting one
w
in the short term, can you bump that capacity to 5 or something? in lieu of trying to figure out why 1 is not acceptable (in case its an algorithm bug or edge case)
m
Sure thing
I bumped to 10 and it has the same error
w
log out of HF. Then in kubernetes delete any outstanding sessions, progreses, virtualmachineclaims
m
Sec
OK, it found two progresses and they have been deleted. I’ve also logged out of everything.
w
Okay, now try again! 😄
m
Errored out immediately
Before it had a spinning icon for a few moments, then errored. This time, straight to error.
w
If you're able, can you share manifests for a) vm template b) environment c) scenario I will try to recreate using latest code on my side
feel free to PM if you don't want those out
m
Sure thing, I’ll send them over in a moment via PM
w
Thanks
f
Have you figured this out? Might also be a problem with the ScheduledEvent when the VMTemplate is changed after creating the Event
m
@faint-optician-47536 Yes, we figured it out. I’m going to be wiping my test environment and re-deploying today to determine if the fix worked without issue. Once I’ve confirmed, I’m going to be submitting updates to the documentation.
👍 1
Open in Slack
PreviousNext
Powered by