Settings View Source Changelog for Oban v2.15

🌟 Looking for changes to Web or Pro? Check the Oban.Pro Changelog or the Oban.Web Changelog. 🌟

🗜️ Notification Compression

Oban uses notifications across most core functionality, from job staging to cancellation. Some notifications, such as gossip, contain massive redundancy that compresses nicely. For example, this table breaks down the compression ratios for a fairly standard gossip payload containing data from ten queues:

Mode	Bytes	% of Original
Original	4720	100%
Gzip	307	7%
Encode 64	412	9%

Minimizing notification payloads is especially important for Postgres because it applies an 8kb limit to all messages. Now all pub/sub notifications are compressed automatically, with a safety mechanism for compatibility with external notifiers, namely Postgres triggers.

🗃️ Query Improvements

There has been an ongoing issue with systems recording a job attempt twice, when it only executed once. While that sounds minor, it could break an entire queue when the attempt exceeded max attempts because it would violate a database constraint.

Apparently, the Postgres planner may choose to generate a plan that executes a nested loop over the LIMITing subquery, causing more UPDATEs than LIMIT. That could cause unexpected updates, including attempts > maxattempts in some cases. The solution is to use a CTE as an "optimization fence" that forces Postgres _not to optimize the query.

We also worked in a few additional query improvements:

• Use an index only scan for job staging to safely handle tables with millions of scheduled jobs.
• Remove unnecessary row locking from staging and pruning queries.

🪶 New Engine Callbacks for SQL Compatibility

We're pleased to share improvements in Oban's SQLite integration. A few SQLite pioneers identified pruning and staging compatibility bugs, and instead of simply patching around the issues with conditional logic, we tackled them with new engine callbacks: stage_jobs/3 and prune_jobs/3. The result is safer, optimized queries for each specific database.

Introducing new engine callbacks with database-specific queries paves the way for working with other databases. There's even an open issue for MySQL support...

v2.15.4 — 2023-08-07

Enhancements

• [Testing] Accept all config options in perform_job/3

  Testing with perform_job/3 requires building a viable config object. Previously, only a few select options like repo were forwarded, which hampered the ability to tweak config for testing.

Bug Fixes

• [PG] Remove monitored processes from listeners list

  The PG notifier monitored listener processes but didn't remove them from the set when a listener exited. Busy systems would gradually accumulate dead pids and kept dispatching to them, bottlnecking the PG process and eventually calling timeouts for new listeners.

v2.15.3 — 2023-08-04

Enhancements

• [Pruner] Prune jobs using the scheduled_at timestamp regardless of state.

  The previous pruning query checked a different timestamp field for each prunable state, e.g. cancelled used cancelled_at. There aren't any indexes for those timestamps, let alone the combination of each state and timestamp, which led to slow pruning queries in larger databases.

  In a database with a mixture of ~1.2m prunable jobs the updated query is 130x faster, reducing the query time from 177ms down to 1.3ms.

• [Lite] Avoid unnecessary transactions during staging and pruning operations

  Contention between SQLite3 transactions causes deadlocks that lead to period errors. Avoiding transactions when there isn't anything to write minimizes contention.

Bug Fixes

• [Foreman] Explicitly pause queues when shutdown begins.

  A call to Producer.shutdown/1 was erroneously removed during the DynamicSupervisor queue refactor.

• [Job] Preserve explicit state set along with scheduled_in time.

  The presence of a scheduled_in timestamp would always set the state to scheduled, even when an explicit state was passed.

v2.15.2 — 2023-06-22

Enhancements

• [Repo] Pass oban: true option to all queries.

  Telemetry options are exposed in Ecto instrumentation, but they aren't obviously available in the opts passed to prepare_query. Now all queries have an oban: true option so users can ignore them in multi-tenancy setups.

• [Engine] Generate a UUID for all Basic and Lite queue instances to aid in identifying orphaned jobs or churning queue producers.

• [Oban] Use Logger.warning/2 and replace deprecated use of :warn level with :warning across all modules.

Bug Fixes

• [Job] Validate changesets during Job.to_map/1 conversion

  The Job.to_map/1 function converts jobs to a map "entry" suitable for use in insert_all. Previously, that function didn't perform any validation and would allow inserting (or attempting to insert) invalid jobs during insert_all. Aside from inconsistency with insert and insert!, insert_all could insert invalid jobs that would never run successfully.

  Now to_map/1 uses apply_action!/2 to apply the changeset with validation and raises an exception identical to insert!, but before calling the database.

• [Notifier] Store PG notifier state in the registry for non-blocking lookup

  To avoid GenServer.call timeouts when the system is under high load we pull the state from the notifier's registry metadata.

• [Migration] Add primary_key explicitly during SQLite3 migrations

  If a user has configured Ecto's :migration_primary_key to something other than bigserial the schema is incompatible with Oban's job schema.

v2.15.1 — 2023-05-11

Enhancements

• [Telemetry] Add [:oban, :stager, :switch] telemetry event and use it for logging changes.

  Aside from an instrumentable event, the new logs are structured for consistent parsing on external aggregators.

• [Job] Remove default priority from job schema to allow changing defaults through the database

Bug Fixes

• [Basic] Restore attempt < max_attempts condition when fetching jobs

  In some situations, a condition to ensure the attempts don't exceed max attempts is still necessary. By checking the attempts outside the CTE we maintain optimal query performance for the large scan that finds jobs, and only apply the check in the small outer query.

• [Pruner] Add missing :interval to Oban.Plugins.Pruner.option/0

v2.15.0 — 2023-04-13

Enhancements

• [Oban] Use DynamicSupervisor to supervise queues for optimal shutdown

  Standard supervisors shut down in a fixed order, which could make shutting down queues with active jobs and a lengthy grace period very slow. This switches to a DynamicSupervisor for queue supervision so queues can shut down simultaneously while still respecting the grace period.

• [Executor] Retry acking infinitely after job execution

  After jobs execute the producer must record their status in the database. Previously, if acking failed due to a connection error after 10 retries it would orphan the job. Now, acking retries infinitely (with backoff) until the function succeeds. The result is stronger execution guarantees with backpressure during periods of database fragility.

• [Oban] Accept a Job struct as well as a job id for cancel_job/1 and retry_job/1

  Now it's possible to write Oban.cancel_job(job) directly, rather than Oban.cancel_job(job.id).

• [Worker] Allow snoozing jobs for zero seconds.

  Returning {:snooze, 0} immediately reschedules a job without any delay.

• [Notifier] Accept arbitrary channel names for notifications, e.g. "my-channel"

• [Telemetry] Add 'detach_default_logger/0' to programmatically disable an attached logger.

• [Testing] Avoid unnecessary query for "happy path" assertion errors in assert_enqueued/2

• [Testing] Inspect charlists as lists in testing assertions

  Args frequently contain lists of integers like [123], which was curiously displayed as '{'.

Bug Fixes

• [Executor] Correctly raise "unknown worker" errors.

  Unknown workers triggered an unknown case error rather than the appropriate "unknown worker" runtime error.

• [Testing] Allow assert_enqueued with a scheduled_at time for available jobs

  The use of Job.new to normalize query fields would change assertions with a "scheduledat" date to _only check scheduled, never "available"

• [Telemetry] Remove :worker from engine and plugin query meta.

  The worker isn't part of any query indexes and prevents optimal index usage.

• [Job] Correct priority type to cover default of 0

For changes prior to v2.15 see the v2.14 docs.