How to find out how much processor resource the hoster has allocated?

Question

linux

CPU usage is 100% on a Debian server even when 5 competing users are doing 1 operation per second, the CPU picture looks something like this: Cpu(s): 82.1%us, 17.9%sy, testing siege -c5 -d1 -r10, me it seems that this is too much load for the Intel® Core (TM) i7 processor declared by the hoster CPU X 980 @ 3.33GHz 1600 Mhz , while there is a lot of RAM left, the server has a bunch of Nginx + Apache + php-fcgi + eAccelerator + memcahe, it is being tested on an empty site managed by livestreet, I look at the CPU load data through top.
How to check how much CPU resources the hosting actually gives out? Or is this load considered normal?

Answer 1

Just on the cheapest VPS from clodo.ru I polled a page for 10kb + 100kb of pictures + nodejs made a request to MongoDB and returned JSON.
Completed 1000 requests at home
ab -n 1000 -c 5 http://*****.net/
Load increased from 1% to 6%

Answer 2

Run some processor benchmark, compare the result with the one found on the Internet (preferably for a processor of the same series, megahertz are different) - find out how much your virtual machine can use a processor.
Compare the results of a test run at different times of the day - you may find out if the hoster is playing with oversell.

Answer 3

I'll add my 5 kopecks ... check, what if your case?

FirstVDS (this is Virtuozzo, a commercial version of OpenVZ), all of a sudden, for no apparent reason, all downloads in php using curl_ ... incredibly slowed down and gave out a load of 100%, after a few hours everything was gone, the reason was not identified.
ps using curl_multi_… eats up memory - it flows very strongly, the bug has been hanging on the tracker for a long time, and periodically it closes and opens

Answer 4

And what do they virtualize? If it's openvz/virtuozo, it's better to change the hoster right away. This is an oversell in 99% of cases, except for the general fs, problems with memory allocation and other unpleasant surprises.
Also, look at what specifically creates a load for you. For example, maybe just some very bad query mysql is trying to digest. =)