[Solved] WP - Not hitting Redis when using Unix Socket

You followed unix socket instructions for redis from Redis - How to install Redis server on Centmin Mod LEMP stack ?

Unfortunately, I now remember that nginx-helper wp plugin itself doesn't support unix sockets as it looks for TCP port so you need to use TCP for Redis Nginx level caching. If you have other web apps supporting Redis unix sockets, you can leverage and use them there.

If anyone is a PHP developer or understands the code Allow use of redis sockets instead of tcp:port · Issue #121 · rtCamp/nginx-helper · GitHub, you could try forking/modifying nginx-helper wp plugin to support unix sockets GitHub - rtCamp/nginx-helper: Nginx Helper for WordPress caching, permalinks & efficient file handling in multisite

run hptweaks.sh is debug mode and copy and paste output to pastebin.com or gist.github.com and post link here

Code (Text):

cd /usr/local/src/centminmod/tools
bash -x hptweaks.sh

 

yeah a redis restart without reboot should do the same thing

 

output for command

Code (Text):

systemd-analyze blame --no-pager

 

contents of /etc/rc.local ?

Code (Text):

cat /etc/rc.local

 

stranged tested it on my centos 7.4 server, despite /etc/rc.local showing 'never' for THP, the actual value is set to always on server reboot

Code (Text):

cat /etc/rc.local               
#!/bin/bash
# THIS FILE IS ADDED FOR COMPATIBILITY PURPOSES
#
# It is highly advisable to create own systemd services or udev rules
# to run scripts during boot instead of using this file.
#
# In contrast to previous versions due to parallel execution during boot
# this script will NOT be run after all other services.
#
# Please note that you must run 'chmod +x /etc/rc.d/rc.local' to ensure
# that this script will be executed during boot.

touch /var/lock/subsys/local
ulimit -n 262144
echo 65536 > /sys/module/nf_conntrack/parameters/hashsize
echo never > /sys/kernel/mm/transparent_hugepage/enabled

on server reboot should be set to never but set to always!

Code (Text):

cat /sys/kernel/mm/transparent_hugepage/enabled
[always] madvise never

but other rc.local setting is set

Code (Text):

cat /sys/module/nf_conntrack/parameters/hashsize
65536

checking to see if tuned is auto overriding the value

Code (Text):

tuned-adm active
Current active profile: virtual-guest

Code (Text):

cat /usr/lib/tuned/virtual-guest/tuned.conf
#
# tuned configuration
#

[main]
summary=Optimize for running inside a virtual guest
include=throughput-performance

[sysctl]
# If a workload mostly uses anonymous memory and it hits this limit, the entire
# working set is buffered for I/O, and any more write buffering would require
# swapping, so it's time to throttle writes until I/O can catch up.  Workloads
# that mostly use file mappings may be able to use even higher values.
#
# The generator of dirty data starts writeback at this percentage (system default
# is 20%)
vm.dirty_ratio = 30

# Filesystem I/O is usually much more efficient than swapping, so try to keep
# swapping low.  It's usually safe to go even lower than this on systems with
# server-grade storage.
vm.swappiness = 30

virtual-guest includes throughput-performance tuned profile too

Code (Text):

cat /usr/lib/tuned/throughput-performance/tuned.conf
#
# tuned configuration
#

[main]
summary=Broadly applicable tuning that provides excellent performance across a variety of common server workloads

[cpu]
governor=performance
energy_perf_bias=performance
min_perf_pct=100

[disk]
readahead=>4096

[sysctl]
# ktune sysctl settings for rhel6 servers, maximizing i/o throughput
#
# Minimal preemption granularity for CPU-bound tasks:
# (default: 1 msec#  (1 + ilog(ncpus)), units: nanoseconds)
kernel.sched_min_granularity_ns = 10000000

# SCHED_OTHER wake-up granularity.
# (default: 1 msec#  (1 + ilog(ncpus)), units: nanoseconds)
#
# This option delays the preemption effects of decoupled workloads
# and reduces their over-scheduling. Synchronous workloads will still
# have immediate wakeup/sleep latencies.
kernel.sched_wakeup_granularity_ns = 15000000

# If a workload mostly uses anonymous memory and it hits this limit, the entire
# working set is buffered for I/O, and any more write buffering would require
# swapping, so it's time to throttle writes until I/O can catch up.  Workloads
# that mostly use file mappings may be able to use even higher values.
#
# The generator of dirty data starts writeback at this percentage (system default
# is 20%)
vm.dirty_ratio = 40

# Start background writeback (via writeback threads) at this percentage (system
# default is 10%)
vm.dirty_background_ratio = 10

# PID allocation wrap value.  When the kernel's next PID value
# reaches this value, it wraps back to a minimum PID value.
# PIDs of value pid_max or larger are not allocated.
#
# A suggested value for pid_max is 1024 * <# of cpu cores/threads in system>
# e.g., a box with 32 cpus, the default of 32768 is reasonable, for 64 cpus,
# 65536, for 4096 cpus, 4194304 (which is the upper limit possible).
#kernel.pid_max = 65536

# The swappiness parameter controls the tendency of the kernel to move
# processes out of physical memory and onto the swap disk.
# 0 tells the kernel to avoid swapping processes out of physical memory
# for as long as possible
# 100 tells the kernel to aggressively swap processes out of physical memory
# and move them to swap cache
vm.swappiness=10

not specifically seeing a setting for transparent_hugepages or is it enabled by default and needs disabling specifically ?

 

strange indeed

what output do you get for

Code (Text):

tuned-adm active

 

will have to investigate more as tried several methods and all on reboot set it back to 'always' instead of 'never'. Only haven't tried override via tuned as want to avoid that as not all systems have tuned enabled.

 

@Nik I just updated centmin mod 123.09beta01's tools/hptweaks.sh routine to add some workaround code to do the following on CentOS 7 systems

1. ensure redis service starts after rc.local
2. ensure 'never' is set in /sys/kernel/mm/transparent_hugepage/enabled on system reboots - did this by adding a custom systemd service called disable-thp that runs on reboot to set it

on most recent centmin mod 123.09beta01 you can update via cmupdate command (on old versions need to use centmin.sh menu option 23 submenu option 2).

Code (Text):

cmupdate
cmdir
tools/hptweaks.sh

then test it out by rebooting server and checking redis logs to see the transparent huge pages warning is present

Code (Text):

reboot

Code (Text):

tail -10 /var/log/redis/redis.log

redis service after rc.local

Code (Text):

systemd-analyze blame --no-pager
         38.514s network.service
          1.438s docker.service
           915ms mariadb.service
           912ms kdump.service
           445ms csf.service
           323ms postfix.service
           290ms lvm2-monitor.service
           287ms php56-php-fpm.service
           287ms php71-php-fpm.service
           284ms php72-php-fpm.service
           273ms php70-php-fpm.service
           272ms nginx.service
           212ms php-fpm.service
           157ms dev-md1.device
           134ms lfd.service
           109ms auditd.service
           107ms pmcd.service
           102ms tuned.service
            60ms systemd-logind.service
            59ms polkit.service
            58ms chronyd.service
            53ms rhel-dmesg.service
            53ms systemd-vconsole-setup.service
            49ms systemd-udev-trigger.service
            49ms iptables.service
            45ms sysstat.service
            43ms systemd-tmpfiles-setup-dev.service
            39ms rhel-readonly.service
            37ms systemd-user-sessions.service
            37ms mdmonitor.service
            25ms sshd.service
            24ms ip6tables.service
            22ms rc-local.service
            21ms redis.service
            21ms systemd-journal-flush.service
            19ms xinetd.service
            17ms systemd-sysctl.service
            15ms memcached.service
            14ms systemd-remount-fs.service
            14ms kmod-static-nodes.service
            13ms systemd-udevd.service
            12ms rhel-import-state.service
            12ms systemd-fsck@dev-md2.service
            10ms systemd-journald.service
            10ms systemd-fsck-root.service
            10ms rsyslog.service
             9ms dev-mqueue.mount
             9ms sys-kernel-debug.mount
             9ms dev-hugepages.mount
             8ms systemd-tmpfiles-setup.service
             7ms systemd-update-utmp-runlevel.service
             6ms systemd-random-seed.service
             6ms systemd-update-utmp.service
             4ms dev-sdb3.swap
             3ms home.mount
             3ms dev-sda3.swap
             3ms systemd-readahead-replay.service
             2ms systemd-readahead-collect.service
             1ms tmp.mount
             1ms systemd-readahead-done.service
           973us sys-kernel-config.mount

 

just leave it in /etc/rc.local as it may get picked up by redis now as the redis service runs after rc.local

 

excellent