NOSQL数据存储＆＃x2B; KV缓存
我正在建立一个系统存储促销活动和一系列促销元数据。
一开始，数据看起来像： pasterion_name，pasterion_content，pasterion_location。 我将使用prosepion_name + proportion_location查询数据
，将来可能会有与新的促销事件相关的额外查询属性，例如pasterion_category。 ..等。我计划将数据存储在诸如DynamoDB之类的NOSQL数据库中，因此我无需使用任何额外的属性迁移表模式。
我有多个数据管道获取促销信息。为了使查询有效，将所有参考字段串联成键（ex：pasterion_location + proportion_category + pasterion_name）和存储他们到另一个KV缓存。还是我应该继续向新字段添加索引（例如dynamodb中的pasterion_category）？
将来最多可能会有10到20个新领域。当检测到新字段时，DynamoDB是否可以将它们存储为第二个索引？或者，只需在后端实现密钥构图逻辑并将其存储在另一个缓存中就很容易。
I am building a system stores the promotion activities and bunch of promotion meta data.
At the beginning, the data looks like:
promotion_name, promotion_content, promotion_location.
And I will query the data with promotion_name + promotion_location
In the future, there could be extra query attributes associated with the new promotion events like promotion_category... etc. I plan to store the data in a NoSQL database like DynamoDB, so I do not need to migrate the table schema with any extra attributes.
I have multiple data pipelines fetching the promotion information. To make the query efficient, would it make sense to concatenate all the reference fields as key (ex: promotion_location + promotion_category + promotion_name) and store them to another K-V cache. Or should I keep adding indexes to the new field (like promotion_category) in DynamoDB?
There're probably going to be at most 10 ~ 20 new fields in the future. Is it possible that DynamoDB can store them as 2nd index when a new field being detected? Or it's easy to just implement key composition logic in backend and store in another cache.
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