lucky-sideburn/kubeinvaders

Seeing Undefined as the Namespace

rohits-splunk opened this issue · 14 comments

Installed my cluster on Digital Ocean and deployed the app by scaling to 5 . I still see "Current Namespace" as Undefined and also i dont see any pods to kill. I followed each and every step to deploy the app but still cant get it to working.

which is the name of namespace you want to stress?

After couple of attempts , i have started seeing namespace1 in the current namespace. Also i am still 0 against all my metrics.

But i am seeing 3 alien pods now,

@rohits-splunk great!! if you vesit the url <your_url>/metrics you don't see anything? Are you using v1.9.5 image?

Yes i am using the v1.9.5 image, i see everything . But the all the metric numbers are 0 . Not sure why. I want to injest these metrics into my monitoring tool, so if they are 0 nothing is gonna show up in the monitoring console.
Screen Shot 2022-10-28 at 2 46 44 PM

chaos_jobs_node_count{node="pool-j7b8g0itq-71lkx"} 2 pods_not_running_on_selected_ns 0 fewer_replicas_seconds 0 log_pod_regex {"pod":".", "namespace":".", "labels":".", "annotations":"."} chaos_jobs_node_count{node="pool-j7b8g0itq-71lky"} 7 chaos_node_jobs_total 9 chaos_jobs_node_count{node="pool-j7b8g0itq-71lkf"} 1 current_chaos_job_pod 0 log_cleaner foobar latest_fewer_replicas_seconds 0 fewer_replicas_time 0

Thank @rohits-splunk! The square are node not pods... Do you see also the aliens? Sorry for this problems :(

No i dont see the aliens.

Ok @rohits-splunk.

As you can see from "developer tools" of the browser the API endpoint for show logs is this: https://kubeinvaders.io/kube/pods?action=list&namespace=namespace1

Please let me know if you see some error. Follow these steps:

  1. Press the button "disable game mode"
  2. Open a terminal with kubectl logs <pod_of_kubeinvader> -n kubeinvaders
  3. Execute from another termina curl "https://<your_kubeinvaders_url>/kube/pods?action=list&namespace=namespace1" -k

Thanks, i am not able to open the link as it says "page cant be loaded".

I followed your steps :
pod_logs.txt

  1. Disabled the game mode
  2. The logs are downloaded ( Attached to this issue )
  3. Cant access this link .

What annotation i should add to the pod get the metrics from this app ? On the OTEL collector config i have added annotation like this :

autodetect.prometheus:true

I think i will have to create another deployment in another namespace to create more aliens to kill . Do you know what method i should use ?

@lucky-sideburn Any update on this ?

Hi @rohits-splunk can you please explain me how OTEL is working? I have to integrate OTEL with an existing Prometheus installation for trying your action.

Thanks

Thanks , I was able to get the data into our monitoring tool. Basically I added appropriate annotations to both OTEL collector pod and Application pod. The internal mechanism took care of scrapping the pods for the metrics.