Shard miturut key sing ana ing meh saben query lan nyebar beban kanthi rata — biasane user_id sing di-hash — banjur rancang resharding lan query antar-shard wiwit dina kapisan. 1 milyar pangguna ora bakal muat utawa mlaku ing siji node; sampeyan misahake row ing pirang-pirang database lan ngarahake saben query menyang shard sing nduweni.
text
┌──────────────┐
Request ─▶ Router/│ shard_of(key)│─▶ Shard 0 (users 0–…)
Proxy └──────────────┘ Shard 1
(Vitess) │ Shard 2
└─────────▶ Shard N ← each = its own primary+replicas
Directory/lookup table (which key → which shard) ── for range or resharding