Problem
The Mandalorian has been away for a while, and he left Baby Yoda (or Grogu) in charge of a little Cassandra key-value store cluster. Poor Baby Yoda doesn't know programming. Because the Mandalorian has been away for a while, compaction has become disabled at every node in his Cassandra key-value store (for a while), while writes and reads have kept on coming in the meantime. Can you explain to Baby Yoda how this disabling of compaction affects:
a. write latency at a single replica node
b. read latency at a single replica node
c. correctness of writes at a single replica node
d. correctness of reads at a single replica node
e. correctness of writes across the cluster (for a given consistency level)
f. correctness of reads across the cluster (for a given consistency level)