Benchmark Linux Servers to Choose the Best One

Linux Servers to Choose the Best One If you want a digital non-public server, or a committed one, you have a whole lot of alternatives. And because the marketplace is flooded and constantly evolves, there’s no “first-class one” for all people – or, as a minimum, no “best one” at everything.

Linux Servers to Choose the Best One Some cloud vendors would possibly provide digital non-public servers with the quickest storage, at the same time as their CPU energy isn’t the greatest. Another issuer might provide the first-class CPU performance in the marketplace, but their storage gadgets are slower. Furthermore, with everyone looking to scouse borrow customers from their competition, the company with sluggish garage may upgrade their devices and turn out to be the first-class in their market, six months later. But, with benchmarks, you could compare extraordinary vendors and select what is satisfactory for you at this second.

A Few Words About Benchmarking

Normally, you’ll adapt benchmarks for your specific use case, along with hardware, and see how an awful lot you may squeeze from your gadgets, their maximum overall performance degrees. However, in this case you’ll run frequent assessments which won’t squeeze out surely every drop of capability. But considering the fact that they’ll run with the identical parameters at the same running gadget, you’ll get actual numbers that you can effortlessly examine and spot which company plays better or worse than some other in equal conditions.

Linux Servers to Choose the Best One Prerequisites

Linux Servers to Choose the Best One Use the modern Ubuntu strong photograph as the running system. At the time of writing, that is 18.04 LTS. It doesn’t rely in case you use another operating gadget later. This will give you a not unusual place to begin. You may also adapt some of the instructions on this academic. If you accomplish that, recall to use the precise identical parameters on all servers to make a accurate comparison.

Linux Servers to Choose the Best One It’s a good idea to run the equal benchmark twice on every server. Some cloud vendors offer inconsistent outcomes. In this example you ought to avoid them altogether because it’s a sign that their guest operating machine isolation isn’t that accurate, or they squeeze too many clients on the equal hardware. Th academic assumes you’re logged in as root. If you’re logged in as a ordinary user, you want to prefix all apt instructions with sudo so that something like apt update && apt set up fio turns into sudo apt update && sudo apt install fio

Linux Servers to Choose the Best One Benchmark Server Storage

apt update && apt install fio

Linux Servers to Choose the Best One If you get a message that fio cannot be determined, it approach you don’t have the universe repository enabled. You can enable it with apt set up software-homes-common && add-apt-repository universe after which repeat the command above to put in fio.

A sequential examine occurs when information is read contiguously. For example, when you examine a 4GB document from begin to complete, it will generally show you the maximum examine speed feasible along with your storage device and the filesystem it’s miles presently the usage of. You can run a sequential read take a look at with:

fio --name=seqread --readwrite=read --direct=1 --ioengine=libaio --bs=1M --size=2000M

If this finishes in much less than twenty seconds (this typically takes place if the garage is on an SSD), you must increase the dimensions of the record being study to get greater accurate effects like the under line of code.

fio --name=seqread --readwrite=read --direct=1 --ioengine=libaio --bs=1M --size=8000M

The most important numbers you should pay attention to, in this result, are READ: bw, highlighted in the following picture.

To test for sequential write speed, run:

fio --name=seqwrite --readwrite=write --direct=1 --ioengine=libaio --bs=1M --size=2000M

To test how your cloud storage performs under the most stressful conditions, run this test:

fio --name=randrw --readwrite=randrw --direct=1 --ioengine=libaio --bs=4k --size=200M --group_reporting --numjobs=8

Just as above, increase --size if the test finishes too fast. In this case, bandwidth is less important – consider it secondary.

First, look at read: IOPS and write: IOPS.

As a real-world example, the server’s storage would be stressed like this on a very busy website with a very large database that constantly has to be read and written to.

Benchmark Server CPU and Memory

Visit Geekbench’s down load page. Copy the hyperlink to the ultra-modern Geekbench archive and paste it to the wget command. For example, on the time of writing this hyperlink is “http://cdn.Geekbench.Com/Geekbench-four.3.3-Linux.Tar.Gz.” The following command will down load Geekbench to your server.



Extract files from the archive.

tar -xzvf *.tar.gz

Change to the extracted directory, which will be equivalent to the program version you found available and was output in the previous command (like in the image above).

cd Geekbench-4.3.3-Linux

At the time of writing the name of the executable file is geekbench4, but this may change in the future. List files in your current directory.


Run the benchmark, replacing the name of the executable, if necessary.


It will take some time until all assessments are finished. At the quit you’ll be presented a link wher eyou can see the consequences.

Benchmark Server’s Network Bandwidth

Install Speedtest client.

apt install speedtest-cli