Redis 8 AGPLv3 Vector JSON TimeSeries 87% Faster 16x Query Docker Setup ไทย

Redis 8 (พ.ค. 2025): Vector/JSON/TimeSeries AGPLv3 เร็วขึ้น 87% Query 16x Redis 8 อัปเกรดใหญ่ Vector search AI JSONPath Time Series IoT Probabilistic structures (Bloom/Cuckoo) Query Engine ACL ใหม่ AGPLv3 Open Source 100%

Redis 8 vs Redis 7 Performance

ฟีเจอร์ใหม่ใช้งานจริง

1. Vector Search (AI/RAG)

# Semantic search
XADD docs * embedding @vector
FT.SEARCH docs "@embedding:[VECSIM 0.8 $query_vec]"

2. JSONPath Query

JSON.SET user:1 $ '{"name": "สมชาย", "age": 30}'
JSON.GET user:1 $.name          # "สมชาย"
JSON.SEARCH users "*.age > 25"  # JSON results

3. Time Series (IoT/Finance)

TS.CREATE temp:room1
TS.ADD temp:room1 * 25.5
TS.RANGE temp:room1 - 24h

Docker Setup 1 คำสั่ง

docker run -d -p 6379:6379 \
  --name redis8 redis:8-alpine
# Stack ครบ (Vector+JSON+TS)
docker run -d -p 6379:6379 redis/redis-stack:8.0

Redis Query Engine (SQL-like)

# Secondary index + Full-text
FT.CREATE idx ON JSON PREFIX 1 users: SCHEMA $.name AS name TEXT_SORT
FT.SEARCH idx "สมชาย | สมศรี" 
# Vector + Fuzzy
FT.SEARCH docs "@category:tech @embedding:[VECSIM 0.7 $vec]"

Probabilistic Structures (Memory 1/10)

Production Use Cases ไทย

🛒 Shopee: Real-time inventory (TimeSeries)
🏪 7-11: Product recommendation (Vector)
📊 SCB: Trading data (10M/sec)
🏥 โรงพยาบาล: Patient monitoring (IoT)
🎮 Garena: Leaderboard (Top-K)

ACL Security ใหม่

ACL SETUSER appuser on >password ~ts:* +ts.add +ts.range @json
ACL SETUSER admin on >adminpass ~* +@all

Performance Tuning

# Redis 8 defaults optimized
vm.overcommit_memory = 1
maxmemory 80%
maxmemory-policy allkeys-lru
io-threads 4

Migration Redis 7→8

✅ In-place upgrade (RDB/AOF)
✅ Vector/JSON backward compatible  
✅ AGPLv3 (check cloud provider)
✅ Stack modules included

Cloud Options ไทย

Benchmark ไทย Workload

E-commerce cart: 2.1M ops/sec (Redis 7: 1.1M)
IoT sensors: 5.2M inserts/sec  
Vector RAG: 12K QPS (cosine sim)