Question-: Which of the following detail can be found using the “cassandra-stress� test with regards to the Java garbage collection? A. Number of garbage collection B. Longest garbage collection in milliseconds C. Total garbage collection in milliseconds D. Size of the garbage collection in milliseconds
Answer: A, B, C,D
Explanation: “Cassandra-stress� tool is also good to find the garbage collection detail for the Data model you have designed. And with this tool you would get to know all the listed option after running the stress testing. A. Number of garbage collection B. Longest garbage collection in milliseconds C. Total garbage collection in milliseconds D. Size of the garbage collection in milliseconds
Admin only
Question-: Which of the following information would be printed when you run the nodetool gcstats subcommand? A. Cluster wide garbage collection info B. Node specific JVM gc collection detail. C. Amount of the time spent for entire cluster for garbage collection D. Total number of gc session across the cluster. E. Total time spent for the garbage collection by individual node.
Answer: B,E
Explanation: Remember that nodetool utility is mainly for individual node and most of the information you would get it is not for entire cluster but rather for individual node. And by running the gcstats command you can get the following detail. - Node specific info for the JVM garbage collection - Measurements which accumulates since previous invocation of gcstats. - And report include the following detail a. Length of time taken for the garbage collection. b. Total amount of memory released by doing the gc on that node c. Total number of gc session
Admin only
Question-: If you see the split brain situation for a particular node in the Cassandra cluster, you can utilize the “nodetool gossipinfo� command, for further analysis? A. True B. False
Answer: A Exp: Yes, using gossipinfo subcommand you can get a more detail about the gossip detail for the particular node and can check whether they are having the same SCHEMA entry or not. Which would be helpful for finding the “split-brain� situation for the node.