We have 12 node Cassandra cluster (3.11.2 version). Out of 12 nodes, one node is getting too much mutation dropped stats in nodetool tpstats
command.
While looking into the logs we are getting frequent StatusLogger.java
messages in the system.log
.
The total number of connections on each individual node is the same. (netstats -na | grep 9041
).
Application is not explicitly writing any data on this particular node and the average load on the system is also very much like other nodes.
What might be the other reasons for having too much mutation dropped message on only one node and not on others?
Why we get StatusLogger.java
messages in the system.log
?
Specifications:-
Xmx/Xms:- 31GB
Average load on each node:- 220GB.
RAM:- 128GB
Thanks,
Sunil