Checking for low disk performance reported by WEB

Problem Description

Troubleshooting of low disk performance reported on the web

solution

Step 1: Enter the device background and View disk io, command: iostat -x 1, as shown in the figure below, the hard disk Latency checked by svctm, if ≥ 25ms, it Description Latency large, r_await, w_await are the Software level IO Latency, and %util is the busyness. If the disk Latency is always high, it Description there is a problem with the disk. Go to step Tue. If the disk Latency Low, data Read/Write, and busyness High, proceed to step Wed.

**Step Two: **Use smartctl -a /dev/sdx x is the drive letter to check the disk smart and View items 5, 197, 198, and 199. If the Latency High and there are values, return the disk to the factory for inspection.
**Step Three: **Stop data Read/Write, then use the command hdparm -t /dev/sdx x is the drive letter, Test Connectivity Wed times, if the performance is lower than 50M/S, you can replace the disk, return the whole machine for repair, and if Test Connectivity Connected, ask R&D to find out Cause the disk is busy Read/Write.
**Step Four: **CPU Usage usage is High, Software check is Error, disk smart Connected, power-on Time exceeds 30,000 hours, and Restart cannot Restore, you can return to the factory to replace the disk.

Original Link

https://support.sangfor.com.cn/cases/list?product_id=156&type=1&category_id=27415&isOpen=true