From 694c87dc9ed4081483983b90757d746edeba0b08 Mon Sep 17 00:00:00 2001
From: peaklabs-dev <122374094+peaklabs-dev@users.noreply.github.com>
Date: Tue, 16 Dec 2025 16:51:59 +0100
Subject: [PATCH] docs: improve and update changelog

- add new v5.x-chore/add-queue-system changes to changelog
- improve wording for a few points
---
 CHANGELOG.md | 14 ++++++++++++--
 1 file changed, 12 insertions(+), 2 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 785c3cd3f..f60c62d07 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -13,10 +13,11 @@ All notable changes to this project will be documented in this file.
 - **v4 to v5 upgrade migration**
   - Added Coolify v4 database as `old_pgsql` connection
   -
+- Worker Servers which replace build servers with servers that can also run jobs (horizon workers) in addition to building docker images
 
 ### Changed
 
-- Upgraded all Composer and Node dependencies, most notably: PHP to 8.5 (previously 8.4), TailwindCSS to v4.0 (previously v3) and Laravel to v12 (previously v10)
+- Upgraded all Composer and Node dependencies and adopted their latest syntax and features, most notably: PHP to 8.5 (previously 8.2), TailwindCSS to v4.0 (previously v3) and Laravel to v12 (previously v10)
 - **Docker:**
   - Upgraded all Docker dependencies, most notably: Postgres to v18 (previously v15) and Redis to v8 (previously v7)
   -
@@ -25,6 +26,11 @@ All notable changes to this project will be documented in this file.
   - Use Redis for sessions and expire inactive sessions after 24h (previously 14 days)
   - Encrypt user sessions data
   - Expire password reset tokens after 10 minutes (previously 60 minutes)
+  - Jobs now wait for all DB transactions to be finished before being dispatched which prevents race conditions
+  - Normal jobs (backups, emails, etc.) and deployment jobs now use separate supervisor configurations and defaults
+  - Horizon workers are now restarted after 500 (job workers) or 300 (deployment workers) jobs or after 1 hour to clean up stale memory and CPU usage
+  - Reduced default queue timeouts from 10h to 60s for jobs and 300s for deployments to prevent stale jobs
+  - Increased `balanceCooldown` from 1s to 2s for jobs to reduce CPU spikes
   - Redirect Laravel logs to `stderr` so they can be viewed in docker logs
   - Configured production logging to rotate automatically and keep only the last 10 days of logs to reduce disk usage
   - Changed production log level from `debug` to `warning` to reduce disk usage and avoid logging sensitive information
@@ -44,6 +50,8 @@ All notable changes to this project will be documented in this file.
 ### Fixed
 
 - `laravel.log` file growing indefinitely and consuming excessive disk space
+- Removed logging of failed jobs into the database as we use Horizon for that and it can cause excessive disk usage in some cases
+- On v4 when changing the maximum concurrent builds setting to more than 4 builds the setting is no longer respected because there is a maximum of 4 horizon workers available by default
 -
 
 ### Security
@@ -70,6 +78,7 @@ All notable changes to this project will be documented in this file.
     - Automatically eager load all relationships to prevent N+1 queries
     - Configure models and enforce morph map for polymorphic relationships
     - Enforce immutable dates globally
+    - Disable queue interruption polling to improve performance
     - Fake sleeps and prevent stray HTTP requests in testing
     - Prevent exception truncation in development
     - Use aggressive Vite prefetching for better performance
@@ -83,13 +92,14 @@ All notable changes to this project will be documented in this file.
   - Cancel in-progress action runs when a new run is triggered
   - Improved `SECURITY.md` formatting and wording and added the support policy for `v5.x`
   - Refactored the GitHub issue templates to use issue types and improved formatting and wording
-  - Moved `README.md` assets into `.github/assets/` to easier exclude them from the core repository code
+  - Moved `README.md` assets into `.github/assets/` to more easily exclude them from the core repository code
   - Removed the `chore-remove-labels-and-assignees-on-close.yml` workflow as labels and assignees are now kept when closing Issues and PRs
 
 ### Refactored
 
 - Completely refactored all database migrations for a cleaner, more consistent and stable database schema
 - Completely refactored all database models
+- Queues are now accessed via an enum instead of hardcoded strings
 
 ## Issues