In Cassandra, why does failure of one node lead to a bunch of nodes fail?

In Lecture 4, second part SQL, it was mentioned that “In Cassandra, the failure of one node leads to a bunch of nodes fail”. Why?

Cassandra uses consistent hashing, basically quord structure and finger table, I don’t see why one node failure causes the failure of a bunch of nodes.
In addition, there are virtual nodes to evenly distribute the load and replication.


Cassandra 在实操过程中因为是 consistent hashing,一旦有failure以后就需要迁移数据,在QPS大的情况下会让本不富裕的机能雪上加霜,造成负载进一步加大,造成 cascading failure.