Changelog

1.0.109-alpha

Changed

• [CONDITIONALLY BREAKING] Timestamps are stored as UTC regardless of JVM's time zone (c10bfc9).
• [CONDITIONALLY BREAKING] JobIdStrategy protocol is introduced to support any job-id type (8b71d06).

Timestamps as UTC

NOTE: This change is breaking only if you have workers or job-enqueueing code running in time zones other than UTC. If unsure, check by logging the output of (java.time.ZoneId/systemDefault).

There is no need to migrate existing jobs if your infrastructure is already running in UTC. The rest of this section only applies if you have workers or job-enqueueing code running in time zones other than UTC.

If you have or enqueue jobs when this version of Proletarian goes to production, they may be processed earlier or later than intended depending on your time zone offset. You will have to decide if this is acceptable for your application. If you need precise migration, you can update existing jobs with SQL:

PostgreSQL:

-- Example for Europe/Oslo (UTC+2 in summer)
-- Adjust the timezone to match your current system timezone
UPDATE proletarian.job
SET process_at = (process_at AT TIME ZONE 'Europe/Oslo') AT TIME ZONE 'UTC';

MySQL:

-- Example for Europe/Oslo (UTC+2 in summer)
-- Adjust the timezone to match your current system timezone
UPDATE proletarian.job
SET process_at = CONVERT_TZ(process_at, 'Europe/Oslo', 'UTC');

If you need help managing the transition to UTC, please open an issue or post a message in #proletarian in the Clojurians Slack community.

JobIdStrategy protocol

Proletarian was inflexible in its handling of jobs ids. Only UUIDs were accepted, and there was no support for database-generated job ids. A UuidSerializer protocol handled the serialization and deserialization of UUIDs (required for MySQL support).

This new change introduces the JobIdStrategy protocol, which makes job id handling user-configurable. Its methods generate-id, encode-id and decode-id together encapsulates job id behavior. Default implementations for UUIDs with PostgreSQL and MySQL are provided.

This is a breaking change for anyone that uses the :proletarian/uuid-serializer config option. This option is likely used only by MySQL users.

To upgrade, replace with :proletarian/job-id-strategy (job-id-strategies/->mysql-uuid-strategy).

1.0.89-alpha - 2024-11-04

Added

• :proletarian/handler-fn-mode option for setting :advanced mode (#16)

1.0.86-alpha - 2024-10-23

Added

• Support for MySQL 8.0.1+ and above
• Documentation updates
• Type hints to avoid reflection warnings

1.0.68-alpha - 2022-05-24

Added

• on-polling-error option (#8)
• failed-job-fn option for handling failed jobs (9d50f2a1)
• Enqueue jobs in the future (#6)

Changed

• Updated dependencies (28e042d3)

1.0.54-alpha - 2021-12-02

Changed

• [BREAKING] Removed proletarian.job/retry-strategy multimethod and added :proletarian/retry-strategy-fn option for proletarian.worker/create-queue-worker (22a86fd8)
• Made logger function 2-arity (1df500e9)
• Updated Postgres JDBC driver (cb8d4427)
• Moved to the MIT license (see https://www.juxt.pro/blog/prefer-mit)

Fixed

• Handle interrupted status also when non-InterruptedException exception is caught (c83dc535)
• Don't stop Queue Worker on SQLTransientException (#5)

Added

• Logging instructions in README (fdcc4e91, dd7156c5)

1.0.41-alpha - 2021-06-10

Added

• Extracted proletarian.worker/process-next-job! function. This function is part of the internal machinery of the Proletarian worker, but is being exposed as a public function for use in testing scenarios and in the REPL. (#2)

1.0.38-alpha - 2021-06-04

Changed

• handle-job! is now a function that must be passed as second argument to create-queue-worker. The proletarian.job/handle-job! multimethod has been replaced by this.

1.0.32-alpha - 2021-03-20

Added

• Lots of documentation updates.

Changed

• The :now-fn option to enqueue! has been removed in favor of a :clock option, mirroring the create-queue-worker options.

1.0.21-alpha - 2021-02-07

Initial public release.