Awesome

ExqBatch

ExqBatch provides a building block to create complex workflows using Exq jobs. A batch monitors a group of Exq jobs and creates callback job when all the jobs are processed.

Example

{:ok, batch} = ExqBatch.new(on_complete: [queue: "default", class: CompletionWorker, args: ["complete"]])
{:ok, batch, jid} = ExqBatch.add(batch, queue: "default", class: Worker, args: [1])
{:ok, batch, jid} = ExqBatch.add(batch, queue: "default", class: Worker, args: [2])
{:ok, _batch} = ExqBatch.create(batch)

Checkout documentation for more information.

Config

config :exq,
  middleware: [
    Exq.Middleware.Stats,
    ExqBatch.Middleware,
    Exq.Middleware.Job,
    Exq.Middleware.Manager,
    Exq.Middleware.Logger
  ]

ExqBatch.Middleware middleware must be added before the Exq.Middleware.Job middleware. The middleware is used to track job life cycle.

config :exq_batch,
  # TTL of a batch job. Under normal condition, batch will get deleted
  # on completion. If a batch failed to complete for any reason, it
  # will get deleted after TTL seconds
  ttl_in_seconds: 60 * 60 * 24 * 30,
  # All the data related to batch will be stored under the prefix in redis
  prefix: "exq_batch",
  # Optional. If present, batch middleware will be enabled only for
  # the given queues. If not present, will be enabled for all queues.
  queues: ["default"]

Caveats

• The completion job will get enqueued only once after all the jobs in a group is either done or dead. A Resurrected job does not belong to the batch and will not lead to creation of completion job.

• All the jobs in a batch must be allowed to be complete (either dead or done). Deleting any jobs while they are in retry queue will cause the batch to get stuck and will expire eventually.