Configuration

Config.yml #

The config.yml configures some global measurement settings that are used when executing the runner.py directly or through cron mode.

Here is an example configuration:

postgresql:
 ...
smtp:
 ...
cluster:
  api_url: http://api.green-coding.internal:9142
  metrics_url: http://metrics.green-coding.internal:9142

machine:
  id: 1
  description: "Development machine for testing"

measurement:
  idle-time-start: 10
  idle-time-end: 5
  flow-process-runtime: 1800
  phase-transition-time: 1
  metric-providers:
    linux:
      cpu.utilization.cgroup.container.provider.CpuUtilizationCgroupContainerProvider:
        resolution: 100
      cpu.energy.RAPL.MSR.system.provider.CpuEnergyRaplMsrSystemProvider:
        resolution: 100
      memory.total.cgroup.container.provider.MemoryTotalCgroupContainerProvider:
        resolution: 100
      cpu.time.cgroup.container.provider.CpuTimeCgroupContainerProvider:
        resolution: 100
  #    psu.energy.ac.xgboost.system.provider.PsuEnergyAcXgboostSystemProvider:
  #      resolution: 100
        # This is a default configuration. Please change this to your system!
  #      CPUChips: 1
  #      HW_CPUFreq: 3100
  #      CPUCores: 28
  #      TDP: 150
  #      HW_MemAmountGB: 16

admin:
  # This address will get an email, when a new project was added through the frontend
  email: [email protected]
  # no_emails: True will suppress all emails. Helpful in development servers
  no_emails: True

The postgresql, smtp and cluster key were already discussed in the installation → part.

The machine key has id and description that are mandatory fields and will be registered in the DB on first run.

We will this only focus on the measurement key:

• idle-time-start [integer]: Seconds to idle containers after orchestrating but before start of measurement

• idle-time-end [integer]: Seconds to idle containers after measurement

• flow-process-runtime [integer]: Max. duration in seconds for how long one flow should take. Timeout-Exception is thrown if exceeded.

• phase-transition-time [integer]: Seconds to idle between phases

• metric-providers:

  • linux/macos/common [string]: Specifies under what system the metric provider can run. Common implies it could run on either.
    • METRIC_PROVIDER_NAME [string]: Key specifies the Metric Provider. Possible Metric Providers →
    • METRIC_PROVIDER_NAME.resolution [integer]: sampling resolution in ms

• client: sleep_time: The seconds the job client should wait before retrying to get another job. Some metric providers have unique configuration params:

• PsuEnergyAcXgboostSystemProvider

  • Please look at the always current documentation here to understand what values to plug in here: XGBoost SPECPower Model documentation

Also note that some providers are deactivated by default, because they either need a additional configuration parameters, extra hardware or a specially configured system.

Once you have set them up you can uncomment the line. In this example for instance the line psu.energy.ac.xgboost.system.provider.PsuEnergyAcXgboostSystemProvider and all the lines directly below it.

admin #

The admin key provides no configuration for the measurements themselves, but rather how the framework behaves.

• email: The email where to error mails and debug mails to
• no_emails: If the framework should send emails at all (including error mails).
  • If you have not SMTP configured you must set this to False