从 S3 到 EC2 请求对象的延迟
从 EC2 实例从 S3 获取对象的延迟是多少?
例如,当通过完整路径请求对象时,该对象的实际数据流开始流回之前有多少毫秒。
- 如果对象存在于 S3 上
- 如果对象不存在并且发送 404
延迟,而不是吞吐量。
What's the latency to GET an object from S3 from an EC2 Instance.
For example, how many ms before the actual data stream for that object starts streaming back when requesting a object by it's full path.
- If the object exists on S3
- If the object does not exist and to send a 404
Latency, not Throughput.
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)
2023 年,在
r6g.medium
实例(不太大也不太小)中测得的 p50(中值)延迟为 10 到 15 毫秒,异常值高达 60 毫秒。 您可以使用此实用程序自行检查:2009 年的原始(过时)答案:
我们通常会遇到从 EC2 到 S3 大约 200-300 毫秒的延迟。 它比您预期的要慢得多。
In 2023, p50 (median) latency measured in an
r6g.medium
instance (not too big, not too small) is 10 to 15ms, with outliers up to 60ms. You can use this utility to check it for yourself:Original (outdated) answer from 2009:
We usually experience latency of around 200-300ms from EC2 to S3. It's much slower than you might expect.
我不认为 AWS 对延迟做出保证,但使用 EC2 中的 S3 肯定会比通过网络更快。 如果您想了解 S3 在实践中的表现如何,请查看 http://www.cloudstatus.com/< /a>. 不幸的是,该网站仅提供吞吐量统计数据,但我认为您可能对此感兴趣。 您可能还对其可用性保证感兴趣:http://aws.amazon.com/s3-sla/
I don't think AWS makes guarantees about latency, but using S3 from EC2 is definitely going to be faster than doing it over the web. If you want to get an idea of how well S3 performs in practice, check out http://www.cloudstatus.com/. Unfortunately, that site only gives throughput statistics, but I figure you may be interested in that anyway. You may also be interested in their availability guarantee: http://aws.amazon.com/s3-sla/