简体繁体 English

Azure Redis缓存中的数据丢失

[英]Data loss in Azure Redis Cache

原文 2016-05-18 04:03:36 9 1 azure/ caching

Following statement suggests that, Azure only provide SLA on Cache endpoint and not on Data. 以下陈述表明，Azure仅在缓存终结点上提供SLA，而不在数据上提供SLA。

SLA only covers connectivity to the Cache endpoints. SLA仅涵盖与缓存端点的连接。 The SLA does not cover protection from data loss. SLA不涵盖防止数据丢失的保护。

My question is, in case of data loss, would all the data be lost from my instance. 我的问题是，在数据丢失的情况下，实例中的所有数据都会丢失吗？ Or partial data loss is also possible. 或部分数据丢失也是可能的。 For instance if I have 2K key-value pairs, is it possible that half of the key value pair are lost. 例如，如果我有2K个键值对，那么可能会丢失一半的键值对。

Why I am curious, because in one application I store some processed result in case (which can be built from data stored in sql). 为什么会感到好奇，因为在一个应用程序中我会存储一些处理后的结果以防万一（可以从sql中存储的数据构建）。 But this data processing is time consuming. 但是这种数据处理很耗时。 Now in case data loss happens, I will have to build this cache again. 现在，万一发生数据丢失，我将不得不再次构建此缓存。 If I can be sure that whenever data is lost, all of my key-value pair will be gone and I can run this process again from scratch. 如果我可以确保每当数据丢失时，我所有的键值对都将消失，那么我可以从头开始再次运行此过程。

And If partial loss is possible, then while accessing individual key value pair, if I don't find it in cache, I will have to build this key-value pair. 而且，如果可能造成部分丢失，那么在访问单个键值对时，如果我在缓存中找不到它，则必须构建此键值对。

1 个解决方案

It sounds like perhaps Redis isn't the right solution for your scenario if you care more about data loss than performance / shared state. 如果您更关心数据丢失而不是性能/共享状态，听起来Redis可能不是适合您的方案的解决方案。 Having said this, you could always look at data persistence for Redis , though this requires use of the Premium tier in Azure. 话虽如此，您始终可以查看Redis的数据持久性，尽管这需要使用Azure中的Premium层。