Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!


does your VPS experience lost time (stalls) - test it
New on LowEndTalk? Please Register and read our Community Rules.

All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.

does your VPS experience lost time (stalls) - test it

wwwcomwwwcom Member
edited May 2016 in General

If you are in a container on an overloaded node try this one liner in your console and watch it for several minutes.
You are looking for more than one second apart in the time on the left.

vmstat 1|while read x;do printf "%.12s $x\n" "$(date +%T.%N)";done

(needs procps and coreutils which you should have, otherwise use apt-get, yum, etc.)

don't be confused by the hundredths, I added that so you can see how "on the nose" vmstat and bash are outputting the line, losing a few tenths or hundredths is fine

reminder: problems usually won't show up immediately, you have to watch for minutes

my own results so far:

OVH kvm - loses time even when container is completely idle, up to 8 seconds!

VMbox openvz - so far so good, sometime loses tenth but never a full second

Ramnode SKVM - smooth as silk, doesn't even lose tenth of a second

about vmstat

the first column "r" is the runqueue, which is number of tasks running or waiting to run - if it is mostly zero, your container is mostly idle

the second column "b" is the blockedqueue, which is stuff that is waiting and cannot be interrupted - it should nearly always be zero or it means you have slow I/O (slow disk)

Thanked by 1StevenN

Comments

  • linuxthefishlinuxthefish Member
    edited May 2016

    Even on a node with 100% CPU usage and fairly high iowait simulated in a few containers I get this:

    09:05:20.954 procs -----------memory---------- ---swap-- -----io---- -system-- ----cpu----
    09:05:20.955 r  b   swpd   free   buff  cache   si   so    bi    bo   in   cs us sy id wa
    09:05:20.957 1  0      0 705816      0  18300    0    0     1     1    0   17  0  0 99  0
    09:05:21.938 0  0      0 705856      0  18364    0    0    64     0    0  117  0  1 99  0
    09:05:22.938 0  0      0 705856      0  18364    0    0     0     0    0   54  1  0 99  0
    09:05:23.938 0  0      0 705852      0  18364    0    0     0     0    0   72  0  0 100  0
    09:05:24.938 0  0      0 705848      0  18364    0    0     0     0    0   64  0  1 99  0
    09:05:25.939 0  0      0 705848      0  18368    0    0     0     0    0  108  0  0 100  0
    09:05:26.939 0  0      0 705852      0  18368    0    0     0     0    0   67  0  0 100  0
    09:05:27.939 0  0      0 705860      0  18368    0    0     0     0    0   56  0  1 99  0
    09:05:28.939 0  0      0 705856      0  18368    0    0     0     0    0   38  0  0 100  0
    09:05:29.939 0  0      0 705860      0  18368    0    0     0     0    0   41  0  0 100  0
    09:05:30.939 0  0      0 705988      0  18368    0    0     0     0    0   77  0  0 100  0
    09:05:31.939 0  0      0 705904      0  18368    0    0     0     0    0   37  0  0 100  0
    09:05:32.940 0  0      0 705880      0  18368    0    0     0     0    0   59  1  1 98  0
    09:05:33.940 0  0      0 706116      0  18368    0    0     0     0    0   69  0  0 100  0
    09:05:34.940 0  0      0 706000      0  18368    0    0     0     0    0   54  0  0 100  0
    09:05:35.940 0  0      0 706000      0  18372    0    0     0     0    0   52  0  0 100  0
    09:05:36.940 0  0      0 705984      0  18372    0    0     0     0    0   66  0  0 100  0
    09:05:37.940 0  0      0 705976      0  18372    0    0     0     0    0   57  0  0 100  0
    09:05:38.941 0  0      0 705980      0  18372    0    0     0     0    0   36  1  0 99  0
    09:05:39.941 0  0      0 705980      0  18372    0    0     0     0    0   47  0  1 99  0
    09:05:40.941 0  0      0 705988      0  18372    0    0     0     0    0   54  0  0 100  0
    09:05:41.941 procs -----------memory---------- ---swap-- -----io---- -system-- ----cpu----
    09:05:41.942 r  b   swpd   free   buff  cache   si   so    bi    bo   in   cs us sy id wa
    09:05:41.943 1  0      0 705988      0  18372    0    0     0     0    0   32  0  0 100  0
    09:05:42.941 0  0      0 705988      0  18372    0    0     0     0    0   54  0  0 100  0
    09:05:43.941 0  0      0 705992      0  18372    0    0     0     0    0   36  0  0 100  0
    09:05:44.942 0  0      0 705980      0  18372    0    0     0     0    0   46  0  0 100  0
    
  • wwwcomwwwcom Member

    @linuxthefish - it should be observed for a few minutes (disk cache writes and all that) but good to see

    is that running from a container or the parent node? because the r (runqueue) is almost always zero which means there are more cpu cores than tasks running

  • FuslFusl Member

    All Vultr locations, looking good

  • wwwcomwwwcom Member

    @Fusl your "r" value is also mostly 0 so be sure to watch it for a few minutes

    (that's a whole bunch of containers!)

Sign In or Register to comment.