Why should you care?

MeanVerse production cannot rely on single mongod — hardware fails during trading hours.

See it live — copy this example

Paste into your MeanVerse project (Angular + Express + MongoDB), then run with ng serve / node / mongosh as noted.

// connect with replica set URI
const uri = 'mongodb://mv-db1:27017,mv-db2:27017,mv-db3:27017/meanverse?replicaSet=rs0';

mongoose.connect(uri, { readPreference: 'primaryPreferred' });

// optional analytics read
const stats = await Transaction.find({ tenantId })
  .read('secondaryPreferred')
  .limit(1000);

What happened?

• replicaSet=rs0 in URI enables failover.
• primaryPreferred writes and strongly consistent reads.
• secondaryPreferred offloads read-heavy reports with slight lag.

Practice next

1. Deploy 3-node replica set in staging.
2. Test failover: step down primary, app reconnects.
3. Use secondary reads only when stale OK.
4. Simulate primary election with rs.stepDown().
5. Document RPO/RTO for MeanVerse DR plan.

Remember

Replica set = HA + optional read scale. Primary for writes and strong reads. MeanVerse Atlas uses managed replica sets.

Banking uptime SLA

Primary disk fails; secondary promoted in 12 seconds.

Outcome: Angular users see brief retry; no data loss on acknowledged writes.