Replica sets na MongoDB suna aiki ta yaya?

Question

Accepted Answer

**Replica set** ƙungiya ne ta survinsa MongoDB da suke kiyaye bayyane gida — suna ba da **babbar samuwa** (sakewa ta atomatik) da **famfo**. Ɗaya ne ɓangarorin **babbar** (yana aiki rubuce), waje duk **gidajiya** (suna karantawa bayyane gida na babbar kuma za su iya yin karatu). Ita ce hanyar mongooDB ta daidaita don aminci.

## Tsarin replica set

```text
PRIMARY (1)   → receives all WRITES; replicates changes to secondaries
SECONDARIES (2+) → copy the primary's data (via the oplog); can serve READS
ARBITER (optional) → votes in elections but holds no data (for odd vote counts)

Write → Primary → replicated to Secondaries (via the operations log "oplog")
```

Babbar yana aiki rubutawa kuma yana karantawa su (ta hanyar **oplog**, littafi na aiki) ga gidajiya, waɗanda suke samuwa. Saitin al'ada yana da membobi 3 don aminci.

## Sakewa ta atomatik (babbar samuwa)

```text
If the PRIMARY fails:
  → the secondaries detect it and hold an ELECTION
  → a secondary is automatically elected as the NEW PRIMARY
  → the application (via the driver) reconnects to the new primary
  → minimal downtime, no manual intervention
→ This automatic failover is the key high-availability benefit.
```

**Sakewa ta atomatik** shine babbar daraja: idan babbar ya kasa, replica set na amsa sabon babbar ta atomatik — mai juyawa katsegida ga gajiya maimakon ganon. (Jijiyar lambar fudu ta kada zai kawar jinjina.)

## Karantawa sikeli da niyoyi na karantawa

```js
// reads can be directed to secondaries to offload the primary
db.collection.find().readPref("secondaryPreferred");
// read preferences: primary (default), secondary, nearest, etc.
```

Gidajiya na iya rubuta karatu (ƙarfafawa ƙarfin karantawa), ƙimar ta **niyoyi na karantawa** — ko da yake karatu gidajiya na iya kasua gajiya (saduwar jiya-jiya).

## Hankalin rubuce (daidaita)

```js
// write concern: how many nodes must acknowledge a write before it's confirmed
db.collection.insertOne(doc, { writeConcern: { w: "majority" } });
// w: 1 (primary only — fast, less safe) vs w: "majority" (safer — survives failover)
```

**Hankalin rubuce** yana ƙarfafawa daidaita: `w: "majority"` yana jira wa gajiya yini a gida (aminci - yana tsira da babbar kasa ba tare da asusun bakin kai) vs `w: 1` (sauri amma rubuce na iya asusun sakewa).

## Me ya sa ya matuwa

Replica sets sune tushe ga gudanar da MongoDB da aminci a cikin aiki, don haka fahimta su shine mahimman ilimin babbar matakin don aiki na MongoDB.

Jakari da basiwar shine **babbar samuwa ta hanyar sakewa ta atomatik**: replica set (ƙungiya ta survinsa kiyaye gida gida, da babbar aiki rubuce kuma gidajiya karantawa) ta atomatik zaɓi sabon babbar idan yanzu ya kasa — mai juyawa katsegida ga gajiya maimakon asusun ganon gaba, wanda ya zama dole ga tsarin da dole su ci gaba (kuma an buƙata don jerin aiki da waje sakin aiki).

Fahimta tsarin (babbar, gidajiya, mabayyannu arbiter, oplog don karantawa) kuma **niyoyi/sakewa** yanameti ya zama tushe.

Sanin cewa gidajiya na iya yin **karatu** (ƙarfafawa ƙarfin karantawa ta hanyar **niyoyi na karantawa**, da kavaye na kasuan gajiya daga saduwar jiya-jiya) yana da daraja don aiki.

Damina mahimmanci shine fahimta **hankalin rubuce** — daidaitar ƙarfafawa waɗa `w: "majority"` tabbatarwa da rubuce yana gida ta gajiya (don haka yana tsira da babbar kasa ba tare da asusun bakin kai — mahimmanci don bayyane muhimmu) katuwa `w: 1` (sauri amma asusun acikin sakewa) — tushen daidaita/aiki jiya.

Tun babbar samuwa kuma daidaita sune dole ga survinsa bayyane, kuma tun replica sets (da sakewa ta atomatik, karantawa sikeli ta gidajiya, kuma hankalin rubuce don daidaita) sune hanyar MongoDB ta tushe don su, fahimta replica sets — tsarin, sakewa ta atomatik, niyoyi na karantawa, kuma musamman hankalin rubuce — mahimmanci ne babbar matakin ilimi don fidlikewar MongoDB da aminci a aiki, labari da zafi da waje, kuma batu da nuna fahimta da mongo ta babbar samuwa kuma daidaita kira da aka buƙata don babbar role gwaji da bayyane.