Jubatus サーバは、メモリ上で機械学習に関するデータを管理しています。 メモリ上でデータを管理するという性質上、Jubatusサーバプロセスの終了とともに Jubatus 上のデータは失われます。
Jubatus では、予期せぬプロセスの終了や誤ったオペレーションに備え、バックアップとリカバリのための機能を提供しています。
現在、Jubatus では、以下の手段を提供しています。
2017-12-03
17.6.30
2017 年 6 月 30 日をもって開発/販売を終了しました。
###* | |
# hogehoge test description | |
# @class Hoge | |
### | |
class Hoge | |
###* | |
# fuga is number | |
# @property {Number} fuga | |
### | |
@fuga |
# Config for Nginx to act as a front-end for Riak | |
# The main goal is to proxy all GETs directly to Riak, and disallow anything else (POST, PUT, etc) | |
# Also, disallow use of the map/reduce query links (i.e. /riak/bucket/key/_,_,_) | |
# Config is in /etc/nginx/sites-available/default or somewhere like that | |
# Set up load-balancing to send requests to all nodes in the Riak cluster | |
# Replace these IPs/ports with the locations of your Riak nodes | |
upstream riak_hosts { | |
server 127.0.0.1:8098; |
Original: "Callbacks are imperative, promises are functional: Node's biggest missed opportunity" by James Coglan
Translated by Yuta Okamoto (@okapies)