Geek Answers Handbook

How data will be consistent on a cassandra cluster

I have doubts when I read the datastax documentation on cassandra record consistency. I have a question about how cassandra will maintain a consistent state in the following scenario:

  • Record Consistency Level = Quorum
  • replication rate = 3

According to the docs, when recording occurs, the node coordinator will send this write request to all the replicas in the cluster. If one replica succeeds and the other fails, the node coordinator will send an error response back to the client, but node -1 has successfully written the data and will not roll back.

In this case

Will a read fix (or hinted-handoff or nodetool repair) replicate inconsistent data from node -1 to node -2 and node -3?

If not, how does Cassandra take care not to copy inconsistent data to other replicas?

Can you clarify my question

0
source share
3 answers

You are absolutely right, read repair or other methods will update node -2 and node -3. This means that even a failed record will eventually update other nodes (if at least one of them was successful). Cassandra has nothing like rollbacks that relational databases have.

+2
source

, - - , , .. , , "", "" ", .

0

Cassandra node (3 ), , , , .

node . node , read repair node, .

0
source

More articles:


All Articles