2
votes

I am Working on HIVE. Where should i create a external table. Inside my local file system OR Inside HDFS file system ?. which is recommended for projects.

1
It's always going to be within HDFS.SMA

1 Answers

2
votes

It has to be within a filesystem accessible by the cluster nodes. While 99% of the times this will mean the cluster file system and 99% of the time this will mean HDFS, this is not strictly true: Hadoop can access other filesystems, trivial examples being S3FileSystem or Azure blobs.

One place you cannot use for external Hive tables is your local filesystem. A local file URI will be meaningless for a remote node. It may work for a pseudo-distributed (single node) 'cluster', but not in real world.