Cassandra Cluster Maintenance Problem

When we encountered our partition is frequently affected latency across our entire database. We were pruned to falling behind on compactions as the problem, right? I told you human beings we always introduce a solution, but the solution has problems Compaction when did we ever had to do compaction with with B plus trees never right? Where Cassandra would compact SS tables on desk for more Performing to reads not onlyWe'd also see cascading latency as they know try to come back. Suffer latency increases resulting in a broader end user impactor at the end of the users feel it and Cluster maintenance task also frequently caused trouble.