内存映射存储引擎
从MongoDb的网站来看,目前唯一支持的存储引擎是Memory-Mapped Storage Engine。我想知道在系统崩溃(即电源故障)的情况下它如何支持原子性?它取决于操作系统吗?有没有关于这个主题的参考?
与以下设计相比,此设计的优缺点是什么:
- couchdb 的仅崩溃设计(仅依赖于附加 b+tree)。
- sqlite 的回滚日志文件
谢谢
From MongoDb's site, the current only support storage engine is Memory-Mapped Storage Engine. I am wondering how it supports atomicity in the presence of system crash(i.e power failure)? Does it depend on the OS? Is there any reference about this topic?
What are the pros and cons of this design comparing with:
- couchdb's crash only design which depends on append only b+tree.
- sqlite's rollback journal file
Thanks
如果你对这篇内容有疑问,欢迎到本站社区发帖提问 参与讨论,获取更多帮助,或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱,如果其他用户或者作者回复了您的评论,将不能在第一时间通知您!
发布评论
评论(2)
这是您正在寻找的页面:
http://www.mongodb.org/显示/DOCS/耐久性+和+修复
This is the page you are looking for:
http://www.mongodb.org/display/DOCS/Durability+and+Repair
要么使用副本集来确保数据到达至少 N 个副本节点,要么对于独立情况,您将等待 MongoDB 1.8 拥有专用日志以确保在崩溃时具有更好的持久性(应该在最新的 1.7.X 中提供)不稳定的释放)。
除此之外:谷歌 - 这个问题已经在各种帖子和博客中进行了深入讨论和记录。不知道为什么不自己先研究一下。
Either you use a replica set for ensure that the data arrived on at least N replica nodes or for the standalone case you'll wait for MongoDB 1.8 having a dedicated journal to ensure better durability across crashes (should be available with the latest 1.7.X unstable release).
And apart from that: Google - this issue has been discussed and documented in depth in various postings and blogs. Not sure why don#t research first yourself.