Geek Answers Handbook

RedShift Node Fault Tolerance

I have a RedShift cluster of 4 nodes.

  • When one of the nodes goes down, will the entire cluster be unavailable?
  • If so, for how long?
  • When the cluster returns - does it return exactly to the same point as before the failure, or can the data be thrown back to the S3 snapshot from a few hours ago?
  • How to simulate this situation to test this scenario yourself?

Thanks a lot!

+4
source share
2 answers

If this is the only node failure, Amazon will launch a new node and a data stream from other nodes (each block is written to two different nodes, if any). In this case, you can expect:

  • node + . 3-4 .
  • 3-4 , , . , .
  • - .

S3. S3 :

  • 8
  • RedShift 5
  • ,
+8

- . 20 , ( "", "" ).

1 "", 1h node.

" " :

A node Amazon Redshift "xxx" 2013-12-18 11:42 UTC. .

- , .

, .

+3

More articles:


All Articles