Question-: Why do you want to check the Heap Dump? A. To find which objects are causing Heap to be filled. B. To find which objects are dirty in the Cassandra database C. To find the which all objects are not being referred in JVM D. To remove the old objects from the Heap
Answer: A
Explanation: Heap Dump is a Java Utility, which would help in finding what all objects are available in the Java Heap currently. You can take Heap Dump from any running Java process. However, Cassandra creates the Heap Dump file only when it about to full the Java Heap area. Once Heap dump is created you can utilize this, to find that what all objects are filling up the Heap area and not getting cleaned.
Question-: You should enable the swappiness on Kernel, so that Cassandra can utilize the virtual memory, when there is not enough memory available? A. True B. False
Answer: False
Explanation: For Cassandra, swapping is very bad event and should be avoided altogether. And to disable the swappiness - Turn off swap for the current kernel process - Remove the swap entries from fstab - Change the swappiness settings
- You can check the current list of swap devices using command “swapon -s� - There is no need to reboot machine when swap is turnoff and that can be done using command “swapoff -a� - However, keep in mind that, once you reboot the node this setting would be reset. - If you want to permanently disable the swap devices then use the fstab file, which are located at /etc/fstab. Comment out any rows that are swap devices.
Question-: Which of the following storage devices should be avoided while setting up Cassandra node? A. SAN Storage B. NAS devices C. NFS D. Direct attached storage
Answer: A, B,C Exp: You must avoid the following persistence storage devices while setting up Cassandra cluster. - SAN storage - NAS devices - NFS You should use the Direct Attached Storage and also, you should consider the SSD. Instead of normal hard disk. SSD has the following properties which can make your storage fast enough.
- It does not have any moving part like Hard Disk. And lesser change of getting damaged. - It has a low latency compare to Hard Disk while reading the data randomly. - As you know that in Commit log data are written sequentially. And SSD are very good for sequential write.