not really a system administrator here but really trying to just set up a server (a rented VDS, really) for some friends.
I recently transferred basically game servers/MySQL/web sites over from one VPS to another - while there hasn't been any issues on the new one I keep seeing my system load spike and take up both processors; previous server system load averaged at about .3-.5. Previous server was on Ubuntu 14, I exported a list of packages I installed from there and apt-get installed them on the new server; I also rsync'd most of the files from the old server over as well (I'm thinking I copied over something bad that's messing with my kernel...)
Anyway, here is the results of my uname -a:
Linux ophq 3.19.0-18-generic #18-Ubuntu SMP Tue May 19 18:31:35 UTC 2015 x86_64 x86_64 x86_64 GNU/Linux
And the results of landscape-sysinfo/logging in screen:
Welcome to Ubuntu 15.04 (GNU/Linux 3.19.0-18-generic x86_64) System load: 2.13 Processes: 11 Usage of /: 22.6% of 196.64GB Users logged in: 1 Memory usage: 32% IP address for eth0: 188.8.131.52 Swap usage: 0%
(currently one game server is in use hence the memory usage - I have to reduce how much RAM is allocated to Minecraft from the default values)
Result of top: http://ericbarber.me/serverproblem/top.png
To add to this - if I hit F and then hit S on 'Process Status' and resort the top lists I have 2 commands listed under 'D'... kworker/u30:0 and kworker/u30:1 which leads me to my kernel assumption...
I'm totally stumped on why load average is so high - I had my users test on both MC and the CS:GO servers and they aren't experiencing any lag - I also tested the web servers and they're delivering pages extremely fast (in comparison to the old server.)
I thought it may be an interrupt issue, so here's the results of cat /proc/interrupts:
Along with this, another question suggested running grep . -r /sys/firmware/acpi/interrupts/ and disabling any values above 0...although all my values are 0 unfortunately.
same url as above serverproblem/interrupts.png
I installed perf and did a quick 30 second report - but I don't understand this output too much:
same url as above serverproblem/perf.png
I'll omit CPU info, but it is an Intel Xeon CPU E5-2690, 2 cores, 2gb RAM, and I believe about 500gb harddrive. My apologies if this is a dumb question or has been asked before - I've been working on this for a few hours now and I'm running into dead-ends with Google past just starting over from scratch...which preferably I would like to avoid.
Apologies on the links..new user limitations.
Edit: To add, the results of mpstat:
Linux 3.19.0-18-generic (ophq) 06/05/2015 _x86_64_ (2 CPU) 02:10:35 PM CPU %usr %nice %sys %iowait %irq %soft %steal %guest %gnice %idle 02:10:35 PM all 7.28 0.00 1.72 47.13 0.00 0.09 0.53 0.00 0.00 43.24