From 72fa2d297f0fd5fd0b4802885dda8fba6acfddc5 Mon Sep 17 00:00:00 2001
From: Saravana Kumar <saravanakumardb1@gmail.com>
Date: Sat, 30 May 2026 09:59:05 +0000
Subject: [PATCH] fix(infra): switch cosmos-emulator from vnext-preview to
 stable :latest

The vnext-preview (Postgres-backed) image returned PGCosmosError plaintext
for cross-partition queryFeed calls, crashing @azure/cosmos at JSON.parse.
:latest is HTTPS-only with a self-signed cert, so consumers are gated by
NODE_TLS_REJECT_UNAUTHORIZED=0 (dev-prototype only). platform-service now
points at the real Azure Cosmos account (per .env), so its dependency on
the local emulator service is removed.
---
 docker-compose.yml | 35 ++++++++++++++++++++++++++++-------
 1 file changed, 28 insertions(+), 7 deletions(-)

diff --git a/docker-compose.yml b/docker-compose.yml
index f0c4f497..1817618d 100644
--- a/docker-compose.yml
+++ b/docker-compose.yml
@@ -34,25 +34,36 @@ services:
     restart: unless-stopped
 
   # ── Azure Cosmos DB Emulator (prototype only) ─────────────────
+  # 2026-05-30: switched off `:vnext-preview` (Postgres-backed experimental
+  # image) because cross-partition `queryFeed` returned plain-text
+  # `PGCosmosError(3, "Database query failed: PostgresError(SqlState(EXX000))...")`
+  # for every query, which crashes the @azure/cosmos SDK at JSON.parse. Affected
+  # every collection — point reads worked, but anything filtering by non-PK
+  # fields failed (login, register, OAuth, feature-flag list, etc.).
+  # `:latest` is the stable Linux port of the Windows Cosmos emulator and
+  # returns proper JSON errors. It is HTTPS-only with a self-signed cert,
+  # so consumers in this compose file are gated by NODE_TLS_REJECT_UNAUTHORIZED=0
+  # (dev-prototype only — never set that in production).
   cosmos-emulator:
-    image: mcr.microsoft.com/cosmosdb/linux/azure-cosmos-emulator:vnext-preview
+    image: mcr.microsoft.com/cosmosdb/linux/azure-cosmos-emulator:latest
     ports:
       - '8081:8081'
       - '1234:1234'
     environment:
-      - PROTOCOL=http
+      - AZURE_COSMOS_EMULATOR_PARTITION_COUNT=10
+      - AZURE_COSMOS_EMULATOR_ENABLE_DATA_PERSISTENCE=false
       - ENABLE_EXPLORER=true
       - GATEWAY_PUBLIC_ENDPOINT=cosmos-emulator
     healthcheck:
       test:
         [
           'CMD-SHELL',
-          'bash -lc ''exec 3<>/dev/tcp/127.0.0.1/8080; printf "GET /ready HTTP/1.1\r\nHost: localhost\r\nConnection: close\r\n\r\n" >&3; grep -q "200 OK" <&3''',
+          'curl -sk --max-time 3 https://127.0.0.1:8081/_explorer/emulator.pem > /dev/null',
         ]
       interval: 10s
       timeout: 5s
-      retries: 12
-      start_period: 20s
+      retries: 30
+      start_period: 60s
     restart: unless-stopped
 
   # ── Loki (Log Aggregation) ────────────────────────────────────
@@ -130,6 +141,12 @@ services:
       - PORT=4003
       # Local/dev convenience: ensure Cosmos DB + containers exist.
       - COSMOS_AUTO_INIT=true
+      # 2026-05-30: switched off the local Cosmos emulator (Postgres-backed
+      # vnext-preview broke `queryFeed` with `PGCosmosError`; stable :latest
+      # crashed under load with a core dump). Pointed at the real Azure
+      # Cosmos DB account (`cosmos-mywisprai`, db `bytelyst`) instead. Values
+      # come from `.env`; the cosmos-emulator service block in this compose
+      # file is no longer needed and platform-service no longer depends on it.
       - PLATFORM_SERVICE_URL=http://platform-service:4003
       - EXTRACTION_SERVICE_URL=http://extraction-service:4005
       - MCP_SERVER_URL=http://mcp-server:4007
@@ -139,8 +156,6 @@ services:
         condition: service_healthy
       azurite:
         condition: service_healthy
-      cosmos-emulator:
-        condition: service_healthy
     labels:
       - 'traefik.enable=true'
       - 'traefik.http.routers.platform.rule=PathPrefix(`/api`) || PathPrefix(`/public`) || PathPrefix(`/health`)'
@@ -164,6 +179,10 @@ services:
     environment:
       - PORT=4005
       - PYTHON_SIDECAR_URL=http://localhost:4006
+      # See cosmos-emulator service block: stable image is HTTPS-only with
+      # a self-signed cert. Dev-prototype only.
+      - COSMOS_ENDPOINT=https://cosmos-emulator:8081
+      - NODE_TLS_REJECT_UNAUTHORIZED=0
     depends_on:
       cosmos-emulator:
         condition: service_healthy
@@ -231,6 +250,8 @@ services:
       - NODE_ENV=development
       - PRODUCT_ID=clawcowork
       - COSMOS_ENDPOINT=https://cosmos-emulator:8081
+      # Stable emulator's self-signed cert — dev-prototype only.
+      - NODE_TLS_REJECT_UNAUTHORIZED=0
       - PLATFORM_SERVICE_URL=http://platform-service:4003
       - EXTRACTION_SERVICE_URL=http://extraction-service:4005
     depends_on: