Ali and so on. CPU 100% failure caused by Windows virtual memory enabled on the server

This morning 11:35~11:40 around, due to load balance in the two cloud server CPU occupied suddenly to 100%, resulting in 5 minutes of the site can not be normal access, please bring trouble, please understand!

(The red curve in the image above indicates CPU usage)

After analysis, we confirmed that the CPU 100% problem is related to enabling Windows virtual memory.

The original two cloud servers were disabled for virtual memory, but yesterday caused the server to restart automatically due to a lack of virtual memory (see Cloud Computing Road-Ali Cloud: Disabling the reboot caused by Windows virtual memory), so Windows virtual memory is enabled. The CPU 100% failure was caused by high concurrency during the peak of today's visit.

Prior to the virtual machine comparison test in Aliyun vs Azure, it was found that enabling virtual memory can cause CPU-intensive problems, but it is not yet known whether the paging file is caused by disk IO or the Xen virtual machine. Can only wait for Aliyun high-performance disk online, the virtual storage in High-performance disk to verify.

The current temporary workaround: Increase physical memory (avoid a situation where virtual memory is not sufficient), and disable virtual memory (avoid a problem with CPU 100% under high concurrency).

See more highlights of this column: http://www.bianceng.cnhttp://www.bianceng.cn/Servers/cloud-computing/