0

So, I'm designing a distributed system with multiple redis instances to break up a large amount of streaming writes, but finding it difficult to get a clear picture of how things work.

From what I've read, it seems that a properly configured cluster will automatically shard and redirect requests made on the 'wrong instance' ( say key 'A' maps to instance 1 but is set on instance 2, it will be redirected to instance 1 ) Am I correct in assuming this?

If so, what advantages does an extra proxy and/or library cluster support give me over simply just connecting to one redis instance and letting it do all the work of figuring out where the SETS and GETS should be done?

Christopher Reid
  • 4,318
  • 3
  • 35
  • 74

1 Answers1

1

Cluster support on the client means the client learns where the data is stored and remembers it, next time it tries to read or write a key it goes straight to the correct instance, which improves performance.

Its like calling directory enquires first every time you want to call a business vs just knowing the number of the business.

Chris Tanner
  • 1,624
  • 1
  • 12
  • 20