问题
I am not able to find a specific link that explains to me how the meta data is distributed in MAPR(File meta data). When I look at cloudera / hortonworks /apache hadoop I know the meta data is stored in namenode's memory which is then fetched to locate the nodes that holds the blocks. How does it work in MAPR is what I am trying to understand.
Any help would be greatly appreciated.
回答1:
MapR natively implemented a Network File System (NFS) interface to MapR-FS so that any reads and writes from and to a file system, whether it be to a local file system, Network Attached Storage, or a Storage Area Network, can read and write data from and to MapR-FS.
This is also the reason MapR asks for a raw disk during the installation so that, it reformats the disk as MapR-FS.
Just came across this thread :
http://answers.mapr.com/questions/108/how-is-metadata-stored.html
来源：https://stackoverflow.com/questions/30803054/mapr-file-read-and-write-process