I tried to diagrammatically represent my understanding of the architecture of MapR FS in my way. Though the diagram shows the containers are in a particular sequence, it will be in random format in real time scenario. Please suggest whether my understanding is correct or wrong. or else please correct me where I am wrong.
A Storage pool is made by default 3 disks per node by MapR-FS.
Each node can have one or more storage pools.
Data is sharded into 256 MB Chunks
Chunks are written into containers that are typically 10-32 GB
Each disk is logically split into containers.
Each Container is spread over to different disk on different storage pools of different nodes on a different rack.
The combination of these grouped containers is called as volume.
Data written in one container will be replicated to another container existing in another node of another rack which belongs to same category.
The locations of the containers is tracked by CLDB.
C stands for Container