大网数据广告活动
服务器运维问答

Scaled out app service seems to scale poorly. Directs all traffic to an instance with 100% CPU usage

· 52CE 我爱测

I have an app service plan that has scaled out to two instances. I've managed to get the statistics on the two instances and one of them is at 100% CPU, and the other is at 5% CPU. This is an issue because it seems all HTTP reque…

问题原文

原帖链接 · A_toaster · CC BY-SA 4.0

I have an app service plan that has scaled out to two instances. I've managed to get the statistics on the two instances and one of them is at 100% CPU, and the other is at 5% CPU. This is an issue because it seems all HTTP requests are being sent to the one at 100% CPU usage, and thus the web page is loading very slowly.

I have turned off ARR Affinity as per this page

Is there another reason why all API and HTTP requests are being sent to the same instance? What can I do to balance the load fairly between the two instances?

已采纳解答

原帖链接 · A_toaster · CC BY-SA 4.0

So the problem fixed itself when I downscaled from P2v2 to S3. These are both exactly the same price, one is premium tier, the other is standard tier.

All requests are now being balanced evenly and the website response time has dropped from 10-12 seconds to 60ms, as a good website should.

Still not sure why it happened in the first place

摘自公开归档,保留原文语言,版式经整理,外部媒体与邮箱地址已移除。

资料来源与适用范围

资料来源
Server Fault
适用范围
以原帖系统、软件版本与发布日期为准(2018-05-28)
内容语言
中文;命令、路径和错误标识保留技术原文
本地收录日期
2026-11-15
授权说明
CC BY-SA 4.0

正文与所需资源由本站本地存储。查看来源及更新记录 ↗

请按适用版本和实际日志确认原因;本站没有对每一种系统或设备逐项执行文中的操作。