AWS EFS CloudWatch指标totaliobytes/datareadiobytes max stat显示了1.05万的上限

发布于 2025-01-31 02:58:10 字数 1199 浏览 4 评论 0原文

我正在收集使用AWS EFS(Amazon弹性文件系统)的应用程序的吞吐量指标。我的问题是:为什么我的CloudWatch指标表明我的EFS系统上限的最大统计数据是〜105m?

最大统计数据的示例图/datareadiobytes “

这些统计总和”的总和显示得更高(GBS):

总二虫/datareadiobytes的示例图 “总iobyobytes/datareadiobytes的sum

尚不清楚这105m来自哪里,我的应用程序到底是什么文件系统吞吐量每秒。

我已经阅读了EFS官方文档,包括 performance href =“ https://docs.aws.amazon.com/efs/latest/ug/performance-tips.html” rel =“ nofollow noreferrer”>提示,指标解释,但它们似乎都没有提供答案。性能链接确实提到,EFS系统的存储使用量很少,但我的应用程序使用超过200GB的EFS系统有一个1M的阈值(但值得注意的是它是较低的)。

我还尝试配置吞吐量,但似乎没有什么不同。

I'm collecting throughput metrics of my application that uses AWS EFS (Amazon Elastic File System). My question is: why my CloudWatch metrics shows that the max stat of TotalIOBytes/DataReadIOBytes of my EFS system caps at ~1.05M?

Sample graph of the Max stat of TotalIOBytes/DataReadIOBytes
Sample graph of the Max stat of TotalIOBytes/DataReadIOBytes

While the sum of these stats shows much higher (GBs):

Sample graph of the Sum stat of TotalIOBytes/DataReadIOBytes
Sample graph of the Sum stat of TotalIOBytes/DataReadIOBytes

It's not clear where does this 1.05M come from and what exactly is my application's file system throughput per second.

I have read through EFS official docs including performance, tips, limits and metrics explanation but none of them seems to provide an answer. The performance link does mention that there is a 1M threshold (but also worth noting that it is lower-bound) for EFS system that has very little storage usage but my application uses more than 200GB.

I also tried configuring provisioned throughput but it didn't seem to make a difference.

There is a seemingly very similar question asked in here.

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。

扫码二维码加入Web技术交流群

发布评论

需要 登录 才能够评论, 你可以免费 注册 一个本站的账号。

评论(1

水波映月 2025-02-07 02:58:10

这些统计数据的总和更高(GB)

该指标没有报告吞吐量(每秒字节),但它是报告字节。
来自文档

总和统计量是与所有文件系统操作关联的字节总数。

使用总和统计量,图上的每个点显示了一定时间段内的字节量(在图表中定义,但在屏幕截图中不可见)。而最大统计量是此期间最大操作的大小。

示例情况:

  • 1 Mbps吞吐量
  • 将仪表板设置的期限限制为1小时,
  • 应用程序几乎连续达到1小时的限制

在这种情况下,

  • ,我们可能会看到最大统计数据绘制了1MB:该期间中最大的操作不能超过吞吐量限制
  • 绘制小于3600 Mbps的数字(1小时内3600秒)

While the sum of these stats shows much higher (GBs)

this metric is not reporting throughput (Bytes per seconds), it's reporting Bytes.
from the documentation linked in the question

The Sum statistic is the total number of bytes associated with all file system operations.

Using the Sum statistic, each point on the graph is showing the amount of bytes for a certain period of time (defined in your chart, but not visible in the screenshots). While the Maximum statistic is the size of the largest operation during such period.

Example situation:

  • 1 MBps throughput limit
  • the period in the dashboard set to 1 hour
  • the application almost hitting the limit for 1 hour consecutively

In this case, we may see

  • the Maximum stat plotting a 1MB: biggest operation in the period cannot exceed the throughput limit
  • the Sum stat plotting a number less than 3600 MBps (3600 seconds in 1 hour period)
~没有更多了~
我们使用 Cookies 和其他技术来定制您的体验包括您的登录状态等。通过阅读我们的 隐私政策 了解更多相关信息。 单击 接受 或继续使用网站,即表示您同意使用 Cookies 和您的相关数据。
原文