Merge pull request #2779 from appwrite/add-edge-resilience-changelog

Add edge resilience changelog

Author: adityaoberai

src/routes/changelog/(entries)/2026-02-20-1.markdoc

@@ -1,21 +1,21 @@
 ---
 layout: changelog
-title: "Free Plan Update: Automatic Pausing of Inactive Projects"
+title: "Free plan update: Automatic pausing of inactive projects"
 date: 2026-02-20
 ---
 
 To reduce idle infrastructure costs and keep the Free plan sustainable at scale, projects on the Free plan with no development activity for 7 consecutive days will be automatically paused.
 
-**What "paused" means**
+**What "paused" means**  
 
 - Data is not deleted
 - The project is not accessible to users while paused
 - No background processes run (backups, cron jobs, scheduled functions)
 - Projects can be reactivated anytime from the Appwrite Console and become active immediately
 
-**Why this change**
+**Why this change**  
 Running cloud infrastructure incurs ongoing costs, even for inactive projects. Automatically pausing idle projects ensures resources are prioritized for developers actively building, while keeping the Free plan sustainable long term.
 
-**Important**
+**Important**  
 Projects that require continuous availability should be on a paid plan. Free plan projects remain active as long as they show development activity in the Console.
 We will continue evolving the Free plan to keep it generous, sustainable, and aligned with active builders.

src/routes/changelog/(entries)/2026-03-02.markdoc

@@ -0,0 +1,15 @@
+---
+layout: changelog
+title: Edge runtimes just got more resilient
+date: 2026-03-02
+---
+
+We've rolled out infrastructure improvements to strengthen runtime stability across regions, including NYC, after identifying DNS-related instability under load.
+
+What's improved:
+
+- DNS resilience under network partitions
+- More powerful instance types and tuning of runtime resource allocations improve performance alongside noisy neighbours
+- Improvements to runtime monitoring and alerting
+
+These changes reduce DNS-related latency, improve cold start reliability, and minimize blast radius during node-level issues.