问题描述

我有一个带有哈希和范围复杂键的表.
我可以使用 AWS SDK for Java 中的 GetItem 查询项目.GetItem 如果没有找到对象，则返回 null，或者作为 Map 的项目返回 null.
我正在寻找检查对象是否存在的最快方法
我在想也许提供一个 .withAttributesToGet 例如:

I have a table with a hash and range complex key.
I can query an item using GetItem from AWS SDK for Java. The GetItem returns null if it doesn't find the object, or the item as a Map<String, AttributeValue>.
I am looking for the fastest approach to check whether the object does exist
I was thinking maybe supplying a .withAttributesToGet such as:

GetItemResult result =  dbClient.getItem(new GetItemRequest().
    withTableName(TABLE_NAME).
        withKey(new Key(new AttributeValue().withS(hashKey),
                        new AttributeValue().withS(rangeKey))).
        withAttributesToGet(new ArrayList<String>()));
Map<String, AttributeValue> item = result.getItem();
return (item != null);

另外一个优化是不使用SDK JSON解析器，自己解析响应，快速检查item是否返回.

Another optimization is to not use the SDK JSON parser and parse the response myself to quickly check if the item has returned.

谢谢

推荐答案

我认为获取"和检查是否存在之间的速度差异可以忽略不计.您可以继续使用 GetItem 本身.如果项目可能太大，则限制返回的属性.

I think there is negligible difference in speed between "getting" and checking if it exists. You can go ahead and use the GetItem itself. If the item is potentially too large, then limit the attributes being returned.

瓶颈在于延迟到达 Dynaamo DB 服务器 (REST API) 和从索引中获取.因此，获取和检查的速度将相似.确保发出调用的服务器与 Dynamo DB 位于同一区域 - 这对速度的影响最大.

The bottle neck is in latency to reach the Dynaamo DB servers (REST API) and in fetching from the index. So Getting and checking will be similar speed. Ensure that your server issuing the call is in the same region as Dynamo DB - This has max impact on the speed.

这篇关于快速查询包含键的表(DynamoDB 和 Java)的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持跟版网！