Hi, I am looking into some issues we are having with a cassandra cluster. What happens is that a node will suddenly start spending all its time in garbage collection becoming non responsive and dropping in and out of the cluster. A restart will solve the problem. We created a heap dump this time and this seems to indicate that there are a lot of scheduled tasks in the nonPeriodicTasks executor.
In particular there are about 40.000 instances of org.apache.cassandra.io.sstable.format.SSTableReader$InstanceTidier$1 Which seems to be created here https://github.com/apache/cassandra/blob/cassandra-3.11/src/java/org/apache/cassandra/io/sstable/format/SSTableReader.java#L2205 Does anyone have any idea what could be causing this ?
