3 ms·
In addition to what the other respondent said, Hive and Impala make it redundant for most use-cases that could work on Hadoop anyway.
by coolsunglasses 13y ago
In addition to what the other respondent said, Hive and Impala make it redundant for most use-cases that could work on Hadoop anyway.
- cmccabe 13y agoHive still has really high query latency, since it starts MR jobs. So it's not really a replacement for what Drawn to Scale was trying to build. Impala certainly is, though. Impala has the additional advantage that your data does not need to be in HBase, but can be in any format in HDFS.