Question-: Whenever incremental_backup flag Is set as true in Cassandra.yaml file then as on when the memtable is flushed it creates the Snapshot for that flushed data as well. A. True B. False
Answer: A
Explanation: Yes by enabling flag “incremental_backup� to true it would create snapshot files from the memtable whenever it is being flushed. Hence both combined snapshot of Memtable as well as SSTables snapshot would be considered, as an entire backup for the table on that node. Still it is not a full snapshot of the table. Snapshot is only created from the data which is available on that node for that token range.
Question-: Which of the following are the correct way to restore snapshot? A. You would be deleting your current data file and copy the snapshot and incremental files to appropriate data directory. B. You can use sstableloader to load the data in the cluster. C. You can use DSBulk tool to load the data from SSTable D. You would be creating new cluster and using the copy command you load the data in your cluster.
Answer: A,B
Explanation: How to restore the data from the snapshot file as well as incremental file there are two options available as below 1. First you delete the current data files and copy the snapshot and incremental file to the appropriate data directories. Remember in this case table schema must be present in order to use this method. Once you have copied the data Restart and prepare the node. 2. Another option is using the sstableloader: this is good if you are having different size of cluster while restoring your snapshot. However, this approach would create lot of load on the cluster for loading the data. But it will properly analyses the token range on the new cluster and copy the data as per the new token range.
Question-: It is recommended that you truncate the table before restoring it. A. True B. False
Answer: A Exp: Yes, it is recommended that you should truncate the table before initiating a restore. Reason being suppose some data have been deleted(accidently) from the table, which would have Tombstone data with the later write timestamp then the same data in the backed-up snapshot file. And if we restore to without truncating or removing the tombstone data, then database will, delete the data which it should not. Because data deletion was done accidentally and you want to load it from the snapshot. But as soon as it is loaded it was deleted again because of the Tombstone marker on that record.