Linux VPS hard disk and network and other major performance tests

With the continuous improvement of storage media (hard disks, etc.) capacity and cost-effectiveness, the development and maturity of VPS technology and the popularization of prices, the application of VPS is becoming more and more extensive. Many webmaster friends have begun to switch their websites from virtual hosts to VPS. For example, there are several cost-effective US VPSs under $10: BurstNET (commonly known as 84vps), DirectSpace, and BuyVM. Of course, using VPS also requires a certain technical foundation. The following is a comprehensive test data on the performance of VPS's hard disk and network compiled by the author. I hope it will be helpful to all webmaster friends.

• VPS TOP information viewing command: top
• VPS CPU information viewing command: cat /proc/cpuinfo
• Use DD to test the read and write performance of VPS hard disk I/O
• VPS egress and ingress network performance test
• Use UB software to test the comprehensive performance of Linux VPS
• Domestic, foreign and VPS website access speed test

1. View top information
top

The main information parameters include the statistical information area: the first five lines are the overall statistical information of the system. The first line is the task queue information, which is the same as the execution result of the uptime command. Its contents are as follows:
09:14:53 Current time
up 3days, 23:47 System uptime, in the format of day, hour:minute
1 user Number of currently logged in users
load average: 0.00, 0.01, 0.03 System load, that is, the average length of the task queue.
The three values ​​are the average values ​​from 1 minute, 5 minutes, and 15 minutes ago to now.
Others include: CPU usage information, memory usage information, system process information, etc.
2. Check CPU information and performance <br /> First use the command to get your CPU information. You can also check it in the probe, which is more detailed.
cat /proc/cpuinfo

The main performance parameters include: how many cores are given to you and what the frequency of each core is. Generally speaking, single-core high frequency is not limited and has better performance than multi-core low frequency.
3. Hard disk IO performance test <br /> Testing the hard disk IO performance will cause great damage to the hard disk. It is not recommended to try multiple times or for a long time. The command is as follows:
dd if=/dev/zero of=test bs=64k count=4k oflag=dsync

Generally speaking, if the test result exceeds 10M, it will have no impact on normal website construction. If it exceeds 50M, it is in a very powerful state. (Note: This test may cause great damage to the hard disk, please use it at your discretion)
4. Network performance test
The network performance of VPS is mainly divided into two indicators: egress and ingress. The ingress can be obtained using the wget file:
wget http://cachefly.cachefly.net/100mb.test

Look at the download speed. If it is greater than 10M/s, it is probably a 100M port. If it reaches 70M/S, it is probably the G port.
5. UB running score to get comprehensive performance <br /> A Linux performance-specific software provided by foreign countries, which can run the score for you to compare:
wget http://www.CTOHome.com/linux-vps-pack/unixbench.sh;sh ./unixbench.sh;
In terms of the total score, scores below 400 indicate low performance; scores between 600 and 800 indicate a normal VPS level; and scores above 1000 indicate excellent performance.
6. VPS access speed test <br /> This mainly tests the PING value and the TRACERT value to determine the speed. Probably these websites:
Webmaster Station: ping.chinaz.com – ping.chinaz.com/tracert
Kaka.com: www.webkaka.com/ping.aspx – www.webkaka.com/Tracert.aspx
Foreign: www.just-ping.com Software: speedtest.net/mini.php
Summary: The above test is from BUYVM's vps, using CentOS5+Nginx+MySQL+PHP (LNMP) architecture. Generally, the results of tests on different mother machines of the same service provider and the same model of VPS may not be the same, because the performance of different mother machines may also be different; and the same mother chicken, using different architectures, may also get different scores. The same VPS will produce different results in different time periods, with different loads on the hens, or when the chicks are unloaded or fully loaded.
via:http://www.webarch.org/41.html