3 ms·
It's not hard to write Hadoop map/reduce queries in other languages: mrjob for python is particularly nice. It's also not relevant that SpiderMonkey is being r
by lucian1900 14y ago
It's not hard to write Hadoop map/reduce queries in other languages: mrjob for python is particularly nice.
It's also not relevant that SpiderMonkey is being replaced with V8.
Mongo's map/reduce is also just a toy, not at all comparable with Hadoop's and being deprecated in favour of the aggregation framework.
Also, there already are decent alternative map/reduce implementations. Disco is a good example, with a similar design to Hadoop.
- kstirman 14y agoMongoDB's MapReduce implementation is not being deprecated. The primary beneficiary of the V8 implementation is MapReduce, so it should be seen as further investment in this area. You can also run Hadoop MapReduce over data in MongoDB directly (24 - 29): http://www.slideshare.net/spf13/mongodb-and-hadoop http://www.slideshare.net/spf13/mongodb-and-hadoop