I have taken a backup of a keyspace containing 8 tables from a 5 node cluster using snapshot tool. The output backup folder contains 5 sub folders which represents each node's data. Each of these 5 sub folders contains the respective tables (8). This is the output backup structure .
When i am trying to restore using sstableloader, i have to loop through each of the 5 folders and subsequently go through each of the 8 tables in that folder and provide the table's path to sstableloader ( my understanding is sstableloader can only accept each table at time ie, i cannot provide the entire tables as list to a single sstableloader command.)
From my current situation of folder structure, i have to call sstableloader 40 times (8 tables * 5 nodes folders) .
Is it possible to avoid this too much calls? It is creating a perfomance overhead . Can i combine the tables or is it somehow possible to provide a list of table folders to a single sstableloader or what could be way to avoid this perfomance overhead?