Piye cara replikasi bisa maidos lan apa wae strategine?

Question

Accepted Answer

**Replikasi** njaga salinan data ing saperangan server — kanggo **ketersediaan** (selamet saka kegagalan), **penskalaan bacaan** (layani bacaan saka salinan), lan **distribusi geografis**. Strategine beda-beda (primary-replica, multi-primary, sinkron vs asinkron) gawe trade-offs.

## Kenapa nreplikasi

```text
✓ AVAILABILITY/redundancy → if one copy fails, others have the data (no data loss, failover)
✓ READ SCALING → serve reads from replicas (offload the primary) → handle more reads
✓ GEO-DISTRIBUTION → copies near users (lower latency); disaster recovery
→ replication is fundamental to reliable, scalable data systems.
```

## Strategi replikasi

```text
PRIMARY-REPLICA (master-slave) → one PRIMARY handles writes, replicates to READ replicas:
  ✓ simple; great for read-heavy workloads; the common model
  ✗ the primary is a write bottleneck / single point for writes (failover needed)
MULTI-PRIMARY (multi-master) → multiple nodes accept writes:
  ✓ write availability/scaling, geo-writes  ✗ CONFLICT resolution (concurrent writes to
    the same data) → complex
```

## Sinkron versus asinkron

```text
SYNCHRONOUS → wait for replicas to confirm before acknowledging the write:
  ✓ replicas are consistent (no data loss on failover)  ✗ slower (waits); availability impact
ASYNCHRONOUS → acknowledge the write immediately; replicate in the background:
  ✓ fast writes, high availability  ✗ REPLICATION LAG → replicas briefly behind (stale reads;
    possible data loss if the primary fails before replicating) — eventual consistency
→ trade-off: durability/consistency vs performance/availability
```

## Kenapa iki penting

Nangerteni replikasi iku penting amerga iku **teknik fundamental kanggo ketersediaan lan penskalaan bacaan** ing sistem data, dadi pengetahuan desain sistem sing penting.

Replikasi — njaga salinan data ing saperangan server — nduweni tujuan penting: **ketersediaan** (selamet saka kegagalan liwat redundansi lan failover), **penskalaan bacaan** (layani bacaan saka replica kanggo ngeceki primary, penting kanggo workload sing akeh bacaan), lan **distribusi geografis** (salinan cedhak pangguna kanggo latency sing luwih cendhak, plus pemulihan bencana).

Nangerteni tujuan-tujuan iki klarifikasi kenapa replikasi iku fundamental kanggo sistem data sing reliable lan scalable.

Nangerteni **strategine** — **primary-replica** (siji primary kanggo nulis nreplikasi menyang read replica — model umum lan simple sing bagus kanggo workload sing akeh bacaan, nanging primary iku bottleneck tulisan) versus **multi-primary** (akeh node nampa tulisan, mamungkinake penskalaan tulisan lan geo-writes nanging mbutuhake resolusi konflik sing rumit) — kudu ana kanggo milih pendekatan sing pas.

Nangerteni **sinkron versus asinkron** trade-off iku utamane penting: replikasi **sinkron** (ngenteni replica ngonfirmasi sebelum ngakoni) ndadekake replica konsisten lan ngindari kehilangan data nanging luwih alon, saengga replikasi **asinkron** (ngakoni langsung, nreplikasi ing latar) cepet lan tinggal ketersediaan nanging duwe **replication lag** (replica mung sedheng ketinggalan, nyebabake stale read lan kemungkinan kehilangan data yen primary gagal sebelum nreplikasi — eventual consistency).

Trade-off durabilitas/consistency versus performance/availability iki keputusan desain sing penting.

Amarga replikasi iku fundamental kanggo ketersediaan lan penskalaan bacaan ing sistem data lan strategine (primary-replica versus multi-primary) lan trade-off sinkron/asinkron dampake consistency, performance, lan ketersediaan, lan amarga nangerti iki penting kanggo ndesain sistem data, nangerteni replikasi iku ilmu desain sistem sing penting lan praktis — teknik fundamental kanggo ketersediaan lan penskalaan bacaan, yen trade-off strategi lan sinkron/asinkron penting, lan sentral kanggo ndesain sistem data sing reliable lan scalable sing selamet saka kegagalan lan nangani beban bacaan.