we planning deploy cassandra cluster 100 virtual nodes. store maximally 1tb (compressed) data on each node. we're going use (host-) local ssd disks.
infrustructure team recommends using sans(even data) since it's easier them backup data.
- which of following methods recommended?
- using more local disks backup
- using local disks data , san backup
- does backup process have overhead slowdown cassandra's write-heavy workloads?
- overhead of copying data backup disk
- overhead of transferring data through network offsite
if cassandra backups nodetool snaphot, backup hard links of data files. , afaik datastax recommends using ssds data files.
