Cron-like job scheduler for Elixir.
To use Quantum in your project, edit the mix.exs
file and add Quantum to both
1. the list of dependencies:
defp deps do
[{:quantum, ">= 1.8.1"}]
end
2. and the list of applications:
def application do
[applications: [:quantum]]
end
Configure your cronjobs in your config/config.exs
like this:
config :quantum, cron: [
# Every minute
"* * * * *": {"Heartbeat", :send},
# Every 15 minutes
"*/15 * * * *": fn -> System.cmd("rm", ["/tmp/tmp_"]) end,
# Runs on 18, 20, 22, 0, 2, 4, 6:
"0 18-6/2 * * *": fn -> :mnesia.backup('/var/backup/mnesia') end,
# Runs every midnight:
"@daily": &Backup.backup/0
]
or like this:
config :quantum, cron: [
# Every minute
"* * * * *": {MyApp.MyModule, :my_method}
]
or you can provide module as a string:
config :quantum, cron: [
# Every minute
"* * * * *": {"MyApp.MyModule", :my_method}
]
Or even use cron-like format (useful with conform / exrm / edeliver):
config :quantum, cron: [
# Every minute
"* * * * * MyApp.MyModule.my_method"
]
If you want to add jobs on runtime, this is possible too:
Quantum.add_job("1 * * * *", fn -> :ok end)
Job struct:
%Quantum.Job{
name: :job_name, # is set automatically on adding a job
schedule: "1 * * * *", # required
task: {MyApp.MyModule, :my_method}, # required
args: [:a, :b] # optional, default: []
state: :active, # is used for internal purposes
nodes: [:node@host], # default: [node()]
overlap: false, # run even if previous job is still running?, default: true
pid: nil, # PID of last executed task
timezone: :utc # Timezone to run task in, defaults to Quantum default which is UTC
}
You can define named jobs in your config like this:
config :quantum, cron: [
news_letter: [
schedule: "@weekly",
task: "MyApp.NewsLetter.send", # {MyApp.NewsLetter, :send} is supported too
args: [:whatever]
]
]
Possible options:
schedule
cron schedule, ex:"@weekly"
or"1 * * * *"
task
function to be performed, ex:"MyApp.MyModule.my_method"
or{MyApp.MyModule, :my_method}
args
arguments list to be passed totask
nodes
nodes list the task should be run on, default:[node()]
overlap
set to false to prevent next job from being executed if previous job is still running, default:true
It is possible to control the behavior of jobs at runtime. For runtime configuration, job names are not restricted to be atoms. Strings, atoms and charlists are allowed as job names.
Add a named job at runtime:
job = %Quantum.Job{schedule: "* * * * *", task: fn -> IO.puts "tick" end}
Quantum.add_job(:ticker, job)
Deactivate a job, i.e. it will not be performed until job is activated again:
Quantum.deactivate_job(:ticker)
Activate an inactive job:
Quantum.activate_job(:ticker)
Find a job:
Quantum.find_job(:ticker)
# %Quantum.Job{...}
Delete a job:
Quantum.delete_job(:ticker)
# %Quantum.Job{...}
If you need to run a job on a certain node you can define:
config :quantum, cron: [
news_letter: [
# your job config
nodes: [:app1@myhost, "app2@myhost"]
]
]
NOTE If nodes
is not defined current node is used and a job is performed on all nodes.
Please note that Quantum uses UTC timezone and not local timezone by default.
To specify another timezone, add the following timezone
option to your configuration:
config :quantum,
cron: [
# Your cronjobs
],
timezone: "America/Chicago"
Valid options are :utc
or a timezone name such as "America/Chicago"
. A full list of timezone names can be downloaded from https://www.iana.org/time-zones, or at https://en.wikipedia.org/wiki/List_of_tz_database_time_zones.
Timezones can also be configured on a per-job basis. This overrides the default Quantum timezone for a particular job. To set the timezone on a job, use the timezone
key when creating the Quantum.Job
structure.
%Quantum.Job{
# ...
timezone: "America/New_York"
}
Field | Allowed values |
---|---|
minute | 0-59 |
hour | 0-23 |
day of month | 1-31 |
month | 1-12 (or names) |
day of week | 0-6 (0 is Sunday, or use abbreviated names) |
Names can also be used for the month
and day of week
fields.
Use the first three letters of the particular day or month (case does not matter).
Instead of the first five fields, one of these special strings may be used:
String | Description |
---|---|
@annually |
Run once a year, same as "0 0 1 1 *" or @yearly |
@daily |
Run once a day, same as "0 0 * * *" or @midnight |
@hourly |
Run once an hour, same as "0 * * * *" |
@midnight |
Run once a day, same as "0 0 * * *" or @daily |
@monthly |
Run once a month, same as "0 0 1 * *" |
@reboot |
Run once, at startup |
@weekly |
Run once a week, same as "0 0 * * 0" |
@yearly |
Run once a year, same as "0 0 1 1 *" or @annually |
The default job settings can be configured as shown in the example below. So if you have a lot of jobs and do not want to override the default setting in every job, you can set them globally.
config :quantum,
cron: [
# Your cronjobs
],
default_schedule: "* * * * *",
default_args: ["my api key"],
default_nodes: [:app1@myhost],
default_overlap: false
Sometimes, you may come across GenServer timeout errors esp. when you have
too many jobs or high load. The default GenServer.call
timeout is 5000.
You can override this default by specifying timeout
setting in configuration.
config :quantum,
timeout: 30_000
Or if you wish to wait indefinitely:
config :quantum,
timeout: :infinity
When you have a cluster of nodes, you may not want same jobs to be generated on every single node, e.g. jobs involving db changes.
In this case, you may choose to run Quantum as a global process, thus preventing same job being run multiple times because of it being generated on multiple nodes. With the following configuration, Quantum will be run as a globally unique process across the cluster.
config :quantum,
cron: [
# Your cronjobs
],
global?: true
This project uses the Collective Code Construction Contract (C4) for all code changes.
"Everyone, without distinction or discrimination, SHALL have an equal right to become a Contributor under the terms of this contract."
- Check for open issues or open a new issue to start a discussion around a problem.
- Issues SHALL be named as "Problem: description of the problem".
- Fork the quantum-elixir repository on GitHub to start making your changes
- If possible, write a test which shows that the problem was solved.
- Send a pull request.
- Pull requests SHALL be named as "Solution: description of your solution"
- Your pull request is merged and you are added to the list of contributors