Learn about Centmin Mod LEMP Stack today
Register Now

Load Testing your Centminmod installation

Discussion in 'Centmin Mod Insights' started by dorobo, Jun 26, 2014.

  1. dorobo

    dorobo Active Member

    420
    104
    43
    Jun 6, 2014
    Ratings:
    +162
    Local Time:
    4:29 AM
    latest
    latest
    I have a WordPress site with WP Super Cache installed which turns the pages into html. Tested it with loader.io, it's free for 1 hostname and you can test up to 10k users per second.

    [​IMG]


    So far it's able to sustain 4k+ users per second without errors or timeouts.

    But if I make it 5k users per second then the errors and timeout starts.

    Perhaps further tuning of sysctl will fix it.
     
    Last edited: Jun 27, 2014
  2. eva2000

    eva2000 Administrator Staff Member

    53,865
    12,160
    113
    May 24, 2014
    Brisbane, Australia
    Ratings:
    +18,712
    Local Time:
    6:29 AM
    Nginx 1.27.x
    MariaDB 10.x/11.4+
    is that with ngx_pagespeed installed but disabled or ngx_pagespeed not installed ?

    play with the following values

    Code:
    worker_processes 2;
    worker_connections  2048;
    i.e.

    Code:
    worker_processes 2;
    worker_connections  8192;
     
  3. pamamolf

    pamamolf Premium Member Premium Member

    4,071
    427
    83
    May 31, 2014
    Ratings:
    +833
    Local Time:
    10:29 PM
    Nginx-1.25.x
    MariaDB 10.3.x
    Can you please share what kind of server it is ? Ram cpu connection speed?

    Also can you post your phpfpm values that you use?
     
  4. dorobo

    dorobo Active Member

    420
    104
    43
    Jun 6, 2014
    Ratings:
    +162
    Local Time:
    4:29 AM
    latest
    latest
    ngx_pagespeed is installed but disabled

    my settings are

    worker_processes auto;
    worker_connections 260000;

    the VPS has 1 CPU so worker_processes being auto defaults to 1.

    It's a Vultr 768MB KVM instance, 1CPU

    Code:
    processor	: 0
    vendor_id	: GenuineIntel
    cpu family	: 6
    model		: 60
    model name	: Vultr Virtual CPU 2
    stepping	: 1
    cpu MHz		: 3392.160
    cache size	: 4096 KB
    physical id	: 0
    siblings	: 1
    core id		: 0
    cpu cores	: 1
    apicid		: 0
    initial apicid	: 0
    fdiv_bug	: no
    hlt_bug		: no
    f00f_bug	: no
    coma_bug	: no
    fpu		: yes
    fpu_exception	: yes
    cpuid level	: 13
    wp		: yes
    flags		: fpu de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx rdtscp lm constant_tsc up unfair_spinlock pni pclmulqdq ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx hypervisor lahf_lm xsaveopt fsgsbase bmi1 hle avx2 smep bmi2 erms invpcid rtm
    bogomips	: 6784.32
    clflush size	: 64
    cache_alignment	: 64
    address sizes	: 40 bits physical, 48 bits virtual
    power management:
     
  5. eva2000

    eva2000 Administrator Staff Member

    53,865
    12,160
    113
    May 24, 2014
    Brisbane, Australia
    Ratings:
    +18,712
    Local Time:
    6:29 AM
    Nginx 1.27.x
    MariaDB 10.x/11.4+
    yeah maybe further tuning of TCP sysctl.conf values might help and with your nginx.conf settings above, make sure you're not running out of memory under load as 768MB might not be enough. If you bump the VPS to a higher memory based one, does the loader.io numbers improve ?

    Again, specific Centmin Mod server optimisation is left to the end user to figure out through monitoring, and trial and error testing ;)
     
  6. dorobo

    dorobo Active Member

    420
    104
    43
    Jun 6, 2014
    Ratings:
    +162
    Local Time:
    4:29 AM
    latest
    latest
    This is top during testing

    [​IMG]

    At 4k+ concurrent users per second. It looks like it doesn't feel a thing. :D
     
  7. rdan

    rdan Well-Known Member

    5,439
    1,399
    113
    May 25, 2014
    Ratings:
    +2,188
    Local Time:
    4:29 AM
    Mainline
    10.2
    And I got this error: