Sign up ×
Unix & Linux Stack Exchange is a question and answer site for users of Linux, FreeBSD and other Un*x-like operating systems. It's 100% free, no registration required.

My Bachelor Final Project is about benchmarking visualization methods with ftp, web and mysql services:

installing Debian on xen (para-visualization) kvm and lxc with equal resources config (equal CPU cores , RAM size and HDD size)

everything is ok but:

Para-ftp Benchmark Jmeter Screenshot

LXC-ftp Benchmark Jmeter Screenshot

I used dedicated Hard Disk with LVM partitions for all lxc , xen and kvm.

My question is: why does this result happen while in other tests with services such as MySQL [for DBMS test] and Apache [for web service], the container based systems go about 2X faster than paravirtualized system?

share|improve this question
You do realize that the interpretation of that data would be the scientific contribution of your work and if you don't do it yourself you're technically a fraud and won't get a degree. People who would hand out Bachelor project like that read SE too, you know. – Bananguin Aug 3 '13 at 19:28
On topic: In general, paravirtualization should always be faster than full virtualization, especially when the VMs are located on the same host. However, depending on the technology you use for storage, a container based system with NAS, maybe even FC, can outperform paravirtualized systems that share the same harddrive(-controller). – Bananguin Aug 3 '13 at 19:34
@user1129682 this my research about performance differences between Visualization methods in this three methods: 1-Full Visualization (kvm) 2-Para Visualization (xen) 3-OS-level Visualization (lxc) in comparing kvm and xen with all tests results are good in comparing xen and lxc in ftp test,containers transferring speed is about half of xen – Sasan torabkheslat Aug 3 '13 at 20:02
compare the lxc containers on the same host vs the containers on different hosts. it may be a scheduling issue – Bananguin Aug 3 '13 at 20:50

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.