Skip to content
sysgeeks

Shard a Live Database Without Downtime, stage 9 of 9: defend it

Why not a distributed database?

Your interviewer: "This is months of custom work. Why not move to a distributed SQL database like CockroachDB, Spanner or Vitess, which shards for you?"

System so far· 7 parts
12345678CLIENTUsersSERVICEApplicationserversDATABASEMonolithPostgresLOG / STREAMWrite audit logWORKERBackfilland catch-upSERVICEConnectionpoolersDATABASEShard hosts

Select a component to see what it is responsible for and which state it owns.

  1. 1Users → Application servers: Requests
  2. 2Application servers → Monolith Postgres: Reads and writes until cutover
  3. 3Application servers → Write audit log: Log every write
  4. 4Backfill and catch-up → Monolith Postgres: Copy existing rows
  5. 5Backfill and catch-up → Write audit log: Replay logged writes
  6. 6Backfill and catch-up → Shard hosts: Write rows by shard
  7. 7Application servers → Connection poolers: Queries by shard
  8. 8Connection poolers → Shard hosts: Schema on host
  • Request / response
  • Asynchronous

What you need to know

0 of 1 checks done

  1. Infrastructure choices are often about risk under a deadline rather than which technology is best in general. Moving to a new database means a full migration plus new operational expertise, while the wraparound deadline approaches.

    A strong answer still concedes what the alternative offers.

  2. Check

    What does a distributed SQL database (CockroachDB, Spanner) genuinely offer over hand-sharding?