You are here
About this site
I am studying Web- and other type of hosting since 1998. Nowadays I use mostly VPSes/cloud and dedicated servers.
This site is dedicated to comparison of VPS plans, providing miscellaneous benchmarks and pieces of advice on how to find, tune and maintain one's VPS.
If you can share an interesting link related to the above, suggest a how-to, offer to publish your review etc., feel free to contact me.
Name: Konstantin Boyandin. I live in Novosibirsk, Russia (time zone Asia/Novosibirsk).
Occupation: system administrator (15+ years of experience), Web programmer (8+ years). I am also a fiction writer, if that's of importance.
How to contact me:
- This site: use contact form
- Email: konstantin [at] vpseer.com or konstantin [at] boyandin.pro
- Jabber/Google Talk: konstantin [at] vpseer.com
- Google +: VPSeer
- Skype: elneirenn
- Twitter: follow @vpseer
- Facebook: VPSeer @Facebook
Every month I buy a number of VPSes for testing purposes. However, I am always glad to test any cloud server /VPS you could provide and/or post your set of benchmarks/run benchmarks for servers of a specific provider.
Testing takes 4 to 72 hours, depending on number of tests you wish to have run, and on hardware capabilities/virtualization type/server load. I run the following tests by default:
- Byte UnixBench
- ioZone Filesystem Benchmarking tool, that tests different types of I/O capabilities
- fio tool I use to get popular IOPS values
- A number of tests from Phoronix Test Suite, namely
If you wish to set up a server for me, I suggest installing CentOS 6.x, 64-bit version, minimal install. I can also install it myself, given you provide me with corresponding tool(s).
Base tests require 128Mb of RAM and 2Gb or more of free disk space available on the server. To run all the tests, 1Gb of RAM is recommended. It doesn't matter, generally, what type of virtualization/technology you use.
Important: in certain cases, say, in OpenVZ-powered VPSes, I can't run certain OS commands — for example, to reset disk caches, in order to receive more realistic results. That will affect certain results (I try to measure IOPS with as large files as possible, thus the more disk you provide, the better — the more precise results we'll see).
Important: please note that the benchmarks are reflecting your servers capabilities just partially. On a real-life server, VPS users will only have a fraction of those high IOPS values (however, for most cases that's quite enough to make customers happy). I can also run any specific type of tests/setting up real-life site and measure its efficiency, in case you need that (that's a paid service, contact me for details).
Live long and prosper!