Mining On A VPS? Important Differences of vServers

Virtual servers (vServers) are the easy and inexpensive way to your own server. Several customers share one physical server – and this is where the problems arise.

Unfortunately, the information provided by the providers on CPU cores, main memory and hard disk is often not meaningful. A vServer with e.g. 4 virtual cores can be very slow in practice if too many customers share the same server. With many providers the resources are not guaranteed and so you may have to share the cores with other customers. The same problems can occur with RAM. If you find a server provider where crypto mining is allowed, you may face the problem that the mining doesn’t work well due to too many people sharing the machine.

For this reason, I have independently tested various vServer offerings and performed standardized benchmarks on them. Tested were the computing power of the CPU, the speed of the RAM and the speed of the hard disk. The fastest offer per category gets 100%. The total value is an average of these three categories.

The performance of different vServer offerings varies massively. Some vendors unfortunately advertise with big promises, e.g. with many cores & a lot of RAM, but the reality is then disappointing. Is it also still possible to mine from your home? Find the answer at Thewowstyle.com.

Decisive for the performance of a vServer is

• Number of vServers per physical server: Some vendors run too many vServers per physical server or single instances consume too many resources. This massively affects the performance of the own server. In spite of supposedly many virtual cores and much RAM, the vServer is then very slow. This problem occurs especially with low-cost vendors who are forced to put too many customers on the same server. In the end, a more expensive package with less cores / RAM can be much faster than the package from a low-cost provider with allegedly many cores / a lot of RAM.
• Server hardware: Your own vServer can only be as fast as the underlying server. The hardware equipment of some physical servers is outdated, accordingly slow can be the CPU, the main memory or the hard disk. For cost reasons, some providers unfortunately also use unsuitable hardware, e.g. cheaper desktop components instead of server components.
• Virtualization software: Which software is used for the virtualization environment has a significant influence on the performance of the vServers. Special server CPUs often offer hardware virtualization support, i.e. the separation of the different vServers can be directly based on the CPU performance. This can achieve up to 99% of the performance of the physical server. However, cheaper CPUs often do not support this, so that some providers resort to software virtualization. In the worst case the vServers reach only 30% of the performance of the physical server.

So far I only tested vendors with a good reputation and no extreme cost dumping at the expense of quality. I would rather leave the cheap providers, who promise a lot, alone, because they are too often a nuisance.

vServer Benchmarks

When I test server performances I usually use the well-known sysbench benchmark suite to test the performance of the CPU, RAM and hard disk. Details about the benchmark can be found on github.

The VPS we use

Personally, we use the vServers from Virmach for some of my own projects and I can absolutely recommend the vServer offerings from Virmach:

• Very high reliability: Up to now I could not detect any malfunctions with my vServers. For crypto mining we use a VPS 200 G8 server from Virmach to run a webserver + database server. So far the accessibility of the application was without any problem.
• Best price on the market: In my opinion Virmach has the best price-performance ratio on the market. The vServers (VPS) have a price guarantee that they are the best offer on the market. The root servers on the other hand have guaranteed CPU and RAM resources.
• Guaranteed resources: The root servers have guaranteed resources (RAM and CPU). Interference from other vServer customers is thus eliminated. This is especially positive for business critical applications.
• Hourly billing: The less expensive vServers (VPS) do not have dedicated CPU resources, but there is no contract period for them and they are billed on an hourly basis. The root servers, on the other hand, unfortunately still have contract periods of between 1 and 12 months.
• Modern hardware: As the benchmark shows, Virmach uses modern (fast) hardware for its servers. Virmach regularly offers new generations for its vServers, in which the latest technology is installed.

Own Server – Do I need it at all?

An own server, whether virtual or not, requires the technical expertise to put it into operation safely. Also a continuous support effort is necessary.

If you only want to run a website, e.g. based on WordPress, I recommend shared web hosting offers. These usually make much less work than an own vServer. In the article “When should you consider VPS Hosting“? they go into more detail about the advantages and disadvantages of an own server and when which decision is the right one.