Part 1 of 8 · 1 min read
Why this page comes now
In this module you will
- Explain why a data set is split into partitions, and how partitioning differs from replication.
- Partition by key range and by hash of the key, and say what each makes easy and hard.
- Compute which partition a key lands on, and the load per partition.
- Spot skew and hot spots in a design, and name a fix for each.
- Combine partitioning with replication and say where each leader lives.
Replication copies the whole data set. It does not help when the data is too big for one machine, or when writes exceed what one leader can take (module 1). Partitioning splits the data so that each machine holds a share. It is the main tool for scaling writes and size. The tier is Core. The numbers are made up to show the idea, unless the text names a book.