Piye cara Redis Cluster ngaktifake skalabilitas?

Question

Accepted Answer

**Redis Cluster** ngaktifake **skalabilitas horizontal** kanthi cara nglompokake data ing sasakbeyan node Redis — ngluwari dataset (lan beban) supaya Redis bisa nangani data lan throughput luwih akeh tinimbang memori server tunggal. Nglompokake data liwat **hash slots** lan menehi saka-awon high availability.

## Sharding liwat hash slots

```text
Redis Cluster divides the keyspace into 16384 HASH SLOTS.
  → each key maps to a slot: slot = CRC16(key) mod 16384
  → slots are distributed across the cluster's primary nodes
  → each node owns a subset of slots (and thus a subset of the data)
→ This spreads data and load across nodes → scales memory and throughput horizontally.
```

Data dipilah ana sasakbeyan node kanthi cara **hash slots** (16384 kabeh): saben key nyusut dadi slot, lan slots disebar ana primary nodes — dadi saben node nyekel bagian data, skalabilitas kapasitas luwih akeh tinimbang siji server.

## High availability kayata gegaman

```text
Each primary node has REPLICAS:
  → if a primary fails, one of its replicas is promoted (automatic failover)
  → the cluster continues serving (HA without separate Sentinel)
→ Redis Cluster combines SHARDING (scale) + REPLICATION (HA).
```

## Clients lan redirection

```text
→ Clients connect to the cluster; a key's request goes to the node owning its slot.
→ If a client hits the wrong node, it gets a MOVED redirect to the right node
  (cluster-aware clients cache the slot map and route directly).
```

## Kontrol multi-key (penting)

```text
⚠️ Multi-key operations only work if all keys are in the SAME slot (same node):
  → operations spanning keys on different nodes are NOT supported directly
  → use HASH TAGS to force related keys into the same slot:
     {user:1000}:profile and {user:1000}:settings → same slot ({...} part is hashed)
→ Design keys so related data that's used together shares a slot.
```

## Apa sebabe penting

Nangertèn Redis Cluster penting kanggo **skalabilitas Redis horizontal luwih akeh tinimbang server tunggal**, dadi katrangan kaya-kaya senior-level kanggo deployment Redis skala gedhe.

Nilai inti yaiku **horizontal scaling**: nalika data utawa throughput ngalampahi apa Redis server tunggal bisa tanggung memori-ne, Cluster **nglumpukake data ana sasakbeyan node** (liwat 16384 **hash slots**, saben key nyusut dadi slot sing disebar ana primary nodes), ngluwari data lan beban kanggo skalabilitas memori lan throughput luwih akeh tinimbang mesin tunggal — penting kanggo sistem skala gedhe.

Nangertèn **mekanisme hash slot** (piye carane key nyusut dadi slots lan slots nyusut dadi nodes) minangka katrangan pokok kanggo piye carane Cluster nglumpukake data.

Nangertèn Cluster menehi **high availability kayata gegaman** (saben primary duwe replica kanthi automatic failover, nggabungake sharding lan replikasi tanpa perlu Sentinel kapisah) penting kanggo kaandalan ing skala.

Nangertèn **client redirection** (MOVED redirects, cluster-aware clients nglumpukake slot map) nerangake piye carane request tekan node sing bener.

Kritisno, nangertèn **kontrol multi-key** — multi-key operations butuh kabeh keys ana slot sing padha, diperangani kanthi **hash tags** (`{user:1000}:...` meksa keys sing gegandhengan ana slot sama) — penting banget amarga iku pembatesan asli sing nepenkowe desain key (data sing digunakake bareng kudu ana sapindhone), lan yen ora nangertèn iku nyebabake kesalahan kanthi multi-key operations ana cluster.

Amarga skalabilitas luwih akeh tinimbang server Redis tunggal butuh kanggo sistem volume dhuwur, lan amarga Redis Cluster (sharding liwat hash slots + high availability kayata gegaman) minangka mekanisme kanggo iku — kanthi pembatesan hash-tag/multi-key penting sing nepenkowe desain — nangertèn Redis Cluster penting katrangan senior-level kanggo skalabilitas Redis, topik sing nunju nangertèn skalabilitas horizontal lan pembatesan praktis (co-lokasi key) sing dadi pertiba-tiba kanthi nglumpukake data Redis ana sasakbeyan nodes.