Scaling databases
too much load
master - slave1, slave2, replicate
master handles write, slave handles read, a lot of read compared to write.
Downsides
- doesn't increase with speed
- replication log
Too much data in one machine
Master - does not fit into one machine, one computer memory
Too much data
Shard
Hash to multiple machines
1-100 101-200 201-300
downsides
- complex queries
- range query - hit all the machines - merge the result -> sort them in the memory - join becomes difficult
replicate and sharding
One setup for replicate
another setup for sharding
No comments:
Post a Comment