Configure your appsettings.yml with these properties
For example
hangfire:
Host: localhost
Port: 5432
Database: postgres
User: postgres
Password: postgres
InMemory: true
enableJobsByFeatureManagementConfig: false
allowedDashboardHost: localhost # you need to set this configuration to be able to access the dashboard from the specified host
invisibilityTimeoutInMinutes: 30
featureManagement: # we can conveniently can use the microsoft feature management system to enable jobs based on configuration
hangfireTestJob: true # if you have enabled the 'enableJobsByFeatureManagementConfig: true' then you can configure here if your jobs should run on execution or not, useful for multiple environments etc.
HangfireConfiguration.cs
[Configuration("hangfire")]
public class HangfireConfiguration
{
public string? Host { get; set; }
public int Port { get; set; }
public string? Database { get; set; }
public string? User { get; set; }
public string? Password { get; set; }
public bool InMemory { get; set; }
public string AllowedDashboardHost { get; set; }
public bool EnableJobsByFeatureManagementConfig { get; set; }
public int? PollingIntervalInMs { get; set; }
public int? WorkerCount { get; set; }
public int InvisibilityTimeoutInMinutes { get; set; } = 30;
public string ConnectionString => $"Host={Host};Port={Port};Database={Database};Username={User};Password={Password};";
}
Configuration in your plugin host
public class SamplePluginHost : IPluginServiceConfiguration
{
public void ConfigureServices(IServiceCollection serviceCollection)
{
// here you can register your recurring jobs or just annotate them with the [RecurringJob] attribute
serviceCollection.ConfigureJobs((jobRegister, _) =>
{
jobRegister.RegisterParameterizedRecurringAsyncJob<HangfireParameterizedTestJob, int>(Cron.Daily(), 1); // every day at 0:00 UTC
jobRegister.RegisterRecurringAsyncJob<HangfireTestJob>(Cron.Hourly(15)); // every hour to minute 15 UTC
});
}
}
Usage
Please take a look at the official Hangfire documentation https://docs.hangfire.io/en/latest/ and best practises https://docs.hangfire.io/en/latest/best-practices.html
Every job does out of the box 10 retries (with increasing wait times) if not configured otherwise.
Auto registration of jobs
You can also define recurring jobs as simple classes and annotate them with the [ReccuringJob] attribute which then auto-registers the job for you at startup
[RecurringJob("0 0 * * *")]
public class HangfireTestJob : AsyncJob
{
public HangfireTestJob(ILogger<HangfireTestJob> logger) : base(logger)
{
}
protected override Task<object> ExecuteJobAsync(CancellationToken cancellationToken)
{
Logger.LogInformation("TEst");
return Task.FromResult<object>(cancellationToken);
}
}
Define jobs
Without parameters
public class HangfireTestJob : AsyncJob
{
public HangfireTestJob(ILogger<HangfireTestJob> logger) : base(logger)
{
}
protected override Task<object> ExecuteJobAsync(CancellationToken cancellationToken)
{
Logger.LogInformation("TEst");
return Task.FromResult<object>(cancellationToken);
}
}
With parameters
Job input parameters and results get serialized to json and are displayed in the hangfire dashboard, be aware of the possibility of leaking sensitive information!
public class MyResult
{
public int JobParameter {get;set;}
}
public class HangfireParameterizedTestJob : ParameterizedAsyncJob<int>
{
public HangfireParameterizedTestJob(ILogger<HangfireParameterizedTestJob> logger) : base(logger)
{
}
protected override Task<MyResult> ExecuteJobAsync(int value, CancellationToken cancellationToken)
{
var result = new MyResult(){ JobParameter = value};
Logger.LogInformation("Test job for {Value}", value);
return result; //
}
}
Fire and forget jobs or enqueue dynamically
You can create new jobs from anywhere in your service, see also https://docs.hangfire.io/en/latest/background-methods/index.html
If you want the job only to be executed as one instance at any given point in time use the
[UniquePerQueueAttribute("high")]
This ensures that there is only one job of the same type/name and method parameters in processing or queued at any given point
JobCleanupExpirationTimeAttribute
With this you can specify how many days the results of a job should be kept until it gets deleted
[JobCleanupExpirationTimeAttribute(14)]
Filters
LogJobExecutionFilter
This filter logs the start and finish of an job execution.
EnabledByFeatureFilter
This filter works in conjunction with the microsoft feature management system. If you would like to be able to enable or disable the execution of your jobs based on configuration this is the right tool for it.
Default Feature management with file based configuration
This works in conjunction with the plugin Dosaic.Plugins.Management.Unleash. This replaces the ability of setting the features via files. Instead one can use the unleash api/service and use feature flags for enabling the jobs. Just make sure that the name of the job matches the name of the feature flag you are creating in unleash.
The dosaic unleash plugin then should do the rest of the heavy lifting.
Since the attribute resolves the feature on before executing the job, changes to the configuration can be done at runtime with a maximal delay based on how often the dosaic unleash plugin refreshes its cache. You can find more information in the documentation of the dosaic unleash plugin.
