LightMySky

Partitioning and Replication for Scale

Beyond one machine, data is split by key across nodes and copied for safety. The partition key decides which queries stay fast and which become scatter-and-gather, and the replication scheme decides what a read may return after a recent write.

No account needed. Progress saves in this browser.

What a learner can do afterwards

  • Choose a partition key for a stated query mix and say which queries it hurts
  • Explain what a hot partition is and how it arises from a skewed key
  • Compare synchronous and asynchronous replication on what a reader can see

1 · Read

Past one machine, you split rows by a partition key and copy each part for safety. The key you choose decides which queries stay fast and which must knock on every door.

Try it together

An orders table keyed by customer keeps one customer's orders together, so that lookup is quick. A report over all recent orders must visit every partition, which is slow scatter and gather. If one seller owns half the rows, its partition runs hot while the rest sit idle.

Copies guard against lost machines. With synchronous replication the write waits for its copies, so later reads see it. With asynchronous replication the write returns fast, but a quick read from a copy can return the older value.

Good to know

Pick the key from your most common query, check the key for skew before you commit, and choose sync or async by asking how stale a read is allowed to be.

Split by the key your queries need, watch for skew, and pick copies that match how fresh your reads must be.

2 · Watch

Take it off screen

Print a worksheetA4 with an answer key page for grown-ups. No screen, no internet.

Where it sits

Then practise

8 questions wait behind this lesson, each with its answer explained. Every answer feeds the sky: stars light as they are learned, and dim when it is time to come back.

Spotted a problem on this page? Tell us
Partitioning and Replication for Scale · Computing, ages 21 to 22 · LightMySky