Upcoming maintenance: Thursday 21 August @ 12:00-14:00 CEST.

Skip to content
Snippets Groups Projects
gifrerenom's avatar
Lluis Gifre Renom authored
- Moved from common/method_wrappers to WebUI
- Generalized definition of dashboards
- Minor corrections in dashboards
- Corrected accounting of "not-found"/"already exists" exceptions in decorator
- Corrected type hints in decorator
- Added deployment of dashboards in deploy script
9c199be4
History
Code owners
Assign users and groups as approvers for specific file changes. Learn more.

Performance Evaluation Method Wrapper

Description:

  • enable prometheus addon:
tfs@tfs-vm:~/tfs-ctrl$ microk8s.enable prometheus
  • wait till prometheus becomes enabled (when enabled, press Ctrl+C):
tfs@tfs-vm:~/tfs-ctrl$ watch -n 1 microk8s.status --wait-ready
  • wait till all pods in the monitoring namespace have STATE=Running and READY=X/X (when done, press Ctrl+C):
tfs@tfs-vm:~/tfs-ctrl$ watch -n 1 kubectl get pods --all-namespaces
  • deploy as:
tfs@tfs-vm:~/tfs-ctrl$ source src/common/method_wrappers/tests/deploy_specs.sh
tfs@tfs-vm:~/tfs-ctrl$ ./deploy.sh
  • expose prometheus and grafana

    • (required) terminal 1 (grafana UI): kubectl port-forward -n monitoring service/grafana --address 0.0.0.0 3001:3000
    • (optional) terminal 2 (prometheus UI): kubectl port-forward -n monitoring service/prometheus-k8s --address 0.0.0.0 9090:9090
    • (optional) terminal 3 (alertmanager UI): kubectl port-forward -n monitoring service/alertmanager-main --address 0.0.0.0 9093:9093
  • if using remote server/VM for running MicroK8s and VSCode, forward ports 3001, 9090, 9093

  • (only used for internal framework debugging) run manual tests over the performance evaluation framework

    • terminal 4:
    export PYTHONPATH=/home/tfs/tfs-ctrl/src
    python -m common.method_wrappers.tests
  • log into grafana:

    • browse: http://127.0.0.1:3000
    • user/pass: admin/admin
    • upload dashboards through "left menu > Dashboards > Manage > Import"
      • upload grafana_prometheus_component_rpc.json
      • upload grafana_prometheus_device_driver.json
      • upload grafana_prometheus_service_handler.json
    • watch in real time the dashboard
  • upload topology through WebUI and navigate

    • should see histogram changing in Grafana

References: