[CASSANDRA-1197] Clean up bmt so that bulk loading cassandra is more approachable for newcomers

• Type: Task
• Status: Resolved
• Priority: Low
• Resolution: Duplicate
• Assignee: Unassigned Assign to me
• Reporter: Jeremy Hanna
• Created: 16/Jun/10 17:34
• Updated: 16/Apr/19 09:33
• Resolved: 05/Aug/10 17:16

Description

Currently the contrib/bmt example is kind of wacky.

In IRC, Jonathan described the current implementation of bmt:
"it's this weird "one row per column" thing, instead of one row per... row"
"you send a row whose columns are serialized rows."

It seems like a bulk loading system should be more usable and understandable by newcomers since that's often one of the first questions when people ask about cassandra - how do I load my existing data into it.

In addition, it would be nice to better document the bmt/bulk loader example and/or create a wiki page that describes in more detail how to do bulk loading with cassandra.

Issue Links

• duplicates: CASSANDRA-1278 Make bulk loading into Cassandra less crappy, more pluggable