2

As you can see in the image shared below, the server load peaked to almost 20 but the CPU, RAM, DISK IO and Network I/O remained well under their full capacity. My question is:

1). What caused the load? 2). Is it ok and normal to ignore these types of high-load spikes?

Load average

1
  • You could try sorting your processes by CPU usage. Feb 21, 2014 at 12:20

1 Answer 1

3

As long as you have 8 cores a load of 16-20 translates to 2-2.5 single-processor style, which is not bad (basically, two cars at the stop light instead of one). You have a process ... ffmpeg maybe? ... that seems to be hogging a single core (not multi-thread happy I'd guess) and probably doing a lot of I/O (corresponding bump in lower left chart)

I would not worry about this, except to wonder why user www-data is able to kick off CPU intensive processes, or if it's causing slowdowns elsewhere.

2
  • You can see a bigger image at i.stack.imgur.com/rNi2i.png , nginx is running under www-data user and is launching ffmpeg - is this not recommended?
    – iTech
    Feb 21, 2014 at 6:08
  • 3
    I can't say, I don't know what your application is or who it's exposed to. I certainly wouldn't want the general public to be able to drive my electric bill up with transcodes unless I understood everything. Feb 21, 2014 at 6:10

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .