I’m going through this course and I had a question regarding the diagram explaining hash partitioning.
My question is: what happens to the data on the node that went down? Is it assumed that it was replicated on other nodes?
Course: https://www.educative.io/collection/10370001/4891237377638400
Lesson: https://www.educative.io/collection/page/10370001/4891237377638400/4623685057511424
I was wondering the same thing. I understand that we can deal with new writes with hashing, but I’m not sure how we handle new reads to entries whose nodes crashed.
Hello,
The large systems use multiple concepts (scalability, availability, reliability, etc.) together. We have explained each of these concepts separately. Here, we are talking about partitioning that helps systems scale. We have discussed different partitioning techniques. The problem you raised can occur in any case and is not specific to hash partitioning only. If a node is dead, the data on that node becomes unavailable. As you mentioned, data replication is the method to deal with data unavailability.