所以这里有一个问题:如果我使用 S3(而不是 RRS)存储,我是否应该备份我的数据,或者 S3 是否已经处理足够好的多站点数据复制冗余?
如果需要备份,我想听听人们迄今为止已经采取了哪些解决方案。
我认为一个简单的方法就是将完整的数据转储到我的本地硬盘并手动复制它。
编辑:目前我将完全脱离亚马逊运行,因此当亚马逊瘫痪时防止数据无法访问将毫无意义。
我唯一担心的是硬件故障会抹去我的数据。
答案1
这取决于你试图缓解哪些风险
- 确保你始终可以访问数据,即使 Amazon 出现故障、停止提供服务等。为此,你需要本地备份。请记住,你将首先因将数据移入 S3 而付费,然后再次因将数据移出而付费。增量备份在这里是一个优势。
- 意外删除数据。您需要备份,不一定是本地备份。它可能只是 S3 中数据的另一个副本。这将花费更多的存储费用,但您无需为将数据移出 S3 付费。无论 Amazon 进行什么数据冗余,都会复制删除。
- 防止硬件故障时数据丢失。亚马逊的做法可能不错。
答案2
虽然情况和你的不一样,但我的想法是一样的。
http://www.dslreports.com/forum/r25854750-Free-Alternatives-to-Dropbox
(If this is a business & "free" is a requirement, or even desire. If this is a business
& you have no control over your website features. If this is a business & you're only
looking into the clouds. ? Thinking your business, your IT, needs to take more
responsibility for itself. Needs to have redundancy in place. Needs ...)
> What about Wuala? 1 GB free and upgrade either..
What about abc, what about xyz, what about 123?
The point is that if one of them, any of them, is all you have to rely on, then that
certainly is not enough.
If you use these types of services as an additional (to more traditional) methods of
doing things, that is fine. But if all you have is xyz, it is bound to fail.
People complained when Amazon had their recent outage. They're a big boy. A little
guy may have said, "enough", packed up shop & turned of the power. Gone (along with
your data & "infrastructure").