Exports all keys in a riak bucket into a newline-delimited file.

export_bucket_keys.rb

# This function uses net::http streaming and the Riak HTTP endpoint
# to export all keys in a riak bucket into a newline-delimited file.
 
require 'json'
require 'net/http'
 
RIAK_ROOT = 'http://localhost:8098'
 
# bucket: The bucket to export
# path: The folder into which to put the output file
# update_per: Console output every this many keys
def export_bucket_keys(bucket, path = '/mnt', update_per = 1000)
  uri = URI("#{RIAK_ROOT}/buckets/#{bucket}/keys?keys=stream")
  count = 0
  File.open("#{path}/#{bucket}_dump.dat", 'w') do |f|
    Net::HTTP.get_response(uri) do |chunk|
      json_blob = ''
      chunk.read_body do |body|
        json_blob += body if body
        if json_blob && json_blob.length > 1
          begin
            JSON.parse(json_blob)['keys'].each do |k|
              f.puts k
              count += 1
              puts "Wrote #{count} keys (last: #{k})" if count % update_per == 0
            end
            json_blob = ''
          rescue JSON::ParserError # This happens when a JSON object is streamed over multiple chunks
          end
        end
      end
    end
  end
end
 
export_bucket_keys('links', '/mnt', 100000) # Generates /mnt/links_dump.dat with console updates every 100k lines