Defines a quantum Scheduler.

When used, the quantum scheduler expects the :otp_app as option. The :otp_app should point to an OTP application that has the quantum runner configuration. For example, the quantum scheduler:

defmodule Khafra.Scheduler do
  use Quantum, otp_app: :khafra_search
end

Could be configured with:

config :khafra_search, Khafra.Scheduler,
  jobs: [
    {"@daily", {Backup, :backup, []}},
  ]

Configuration:

• :clock_broadcaster_name - GenServer name of clock broadcaster
  
  (unstable, may break without major release until declared stable)

• :execution_broadcaster_name - GenServer name of execution broadcaster
  
  (unstable, may break without major release until declared stable)

• :executor_supervisor_name - GenServer name of execution supervisor
  
  (unstable, may break without major release until declared stable)

• :debug_logging - Turn on debug logging

• :jobs - list of cron jobs to execute

• :job_broadcaster_name - GenServer name of job broadcaster
  
  (unstable, may break without major release until declared stable)

• :name - GenServer name of scheduler
  
  (unstable, may break without major release until declared stable)

• :node_selector_broadcaster_name - GenServer name of node selector broadcaster
  
  (unstable, may break without major release until declared stable)

• :overlap - Default overlap of new Job

• :otp_app - Application where scheduler runs

• :run_strategy - Default Run Strategy of new Job

• :schedule - Default schedule of new Job

• :storage - Storage to use for persistence

• :storage_name - GenServer name of storage
  
  (unstable, may break without major release until declared stable)

• :supervisor_module - Module to supervise scheduler
  
  Can be overwritten to supervise processes differently (for example for clustering)
  
  (unstable, may break without major release until declared stable)

• :task_registry_name - GenServer name of task registry
  
  (unstable, may break without major release until declared stable)

• :task_supervisor_name - GenServer name of task supervisor
  
  (unstable, may break without major release until declared stable)

• :timeout - Sometimes, you may come across GenServer timeout errors esp. when you have too many jobs or high load. The default GenServer.call/3 timeout is 5_000.

• :timezone - Default timezone of new Job

Telemetry

• [:quantum, :job, :add]

  • Description: dispatched when a job is added
  • Measurements: %{}
  • Metadata: %{job: Quantum.Job.t(), scheduler: atom()}

• [:quantum, :job, :update]

  • Description: dispatched when a job is updated
  • Measurements: %{}
  • Metadata: %{job: Quantum.Job.t(), scheduler: atom()}

• [:quantum, :job, :delete]

  • Description: dispatched when a job is deleted
  • Measurements: %{}
  • Metadata: %{job: Quantum.Job.t(), scheduler: atom()}

• [:quantum, :job, :start]

  • Description: dispatched on job execution start
  • Measurements: %{system_time: integer()}
  • Metadata: %{telemetry_span_context: term(), job: Quantum.Job.t(), node: Node.t(), scheduler: atom()}

• [:quantum, :job, :stop]

  • Description: dispatched on job execution end
  • Measurements: %{duration: integer()}
  • Metadata: %{telemetry_span_context: term(), job: Quantum.Job.t(), node: Node.t(), scheduler: atom(), result: term()}

• [:quantum, :job, :exception]

  • Description: dispatched on job execution fail
  • Measurements: %{duration: integer()}
  • Metadata: %{telemetry_span_context: term(), job: Quantum.Job.t(), node: Node.t(), scheduler: atom(), kind: :throw | :error | :exit, reason: term(), stacktrace: list()}

Examples

iex(1)> :telemetry_registry.discover_all(:quantum)
:ok
iex(2)> :telemetry_registry.spannable_events()
[{[:quantum, :job], [:start, :stop, :exception]}]
iex(3)> :telemetry_registry.list_events
[
  {[:quantum, :job, :add], Quantum,
   %{
     description: "dispatched when a job is added",
     measurements: "%{}",
     metadata: "%{job: Quantum.Job.t(), scheduler: atom()}"
   }},
  {[:quantum, :job, :delete], Quantum,
   %{
     description: "dispatched when a job is deleted",
     measurements: "%{}",
     metadata: "%{job: Quantum.Job.t(), scheduler: atom()}"
   }},
  {[:quantum, :job, :exception], Quantum,
   %{
     description: "dispatched on job execution fail",
     measurements: "%{duration: integer()}",
     metadata: "%{telemetry_span_context: term(), job: Quantum.Job.t(), node: Node.t(), scheduler: atom(), kind: :throw | :error | :exit, reason: term(), stacktrace: list()}"
   }},
  {[:quantum, :job, :start], Quantum,
   %{
     description: "dispatched on job execution start",
     measurements: "%{system_time: integer()}",
     metadata: "%{telemetry_span_context: term(), job: Quantum.Job.t(), node: Node.t(), scheduler: atom()}"
   }},
  {[:quantum, :job, :stop], Quantum,
   %{
     description: "dispatched on job execution end",
     measurements: "%{duration: integer()}",
     metadata: "%{telemetry_span_context: term(), job: Quantum.Job.t(), node: Node.t(), scheduler: atom(), result: term()}"
   }},
  {[:quantum, :job, :update], Quantum,
   %{
     description: "dispatched when a job is updated",
     measurements: "%{}",
     metadata: "%{job: Quantum.Job.t(), scheduler: atom()}"
   }}
]