Tuning EC2 with an ElasticBeanstalk ebextension

.ebextensions_00_tune_ec2.config_.yaml

packages:
  yum:
    wget: []
    curl: []
commands:
  00_remove_99_swap.conf.bak:
    command: rm -f /etc/sysctl.d/99_swap.conf.bak
    test: test -f /etc/sysctl.d/99_swap.conf.bak
    ignoreErrors: true
  00_remove_99_filesystem.conf.bak:
    command: rm -f /etc/sysctl.d/99_filesystem.conf.bak
    test: test -f /etc/sysctl.d/99_filesystem.conf.bak
    ignoreErrors: true
  00_remove_99_networking.conf.bak:
    command: rm -f /etc/sysctl.d/99_networking.conf.bak
    test: test -f /etc/sysctl.d/99_networking.conf.bak
    ignoreErrors: true
  00_remove_00tune_ec2.sh.bak:
    command: rm -f /opt/elasticbeanstalk/hooks/appdeploy/pre/00tune_ec2.sh.bak
    test: test -f /opt/elasticbeanstalk/hooks/appdeploy/pre/00tune_ec2.sh.bak
    ignoreErrors: true
files:
  "/opt/elasticbeanstalk/hooks/appdeploy/pre/00tune_ec2.sh":
    mode: "000755"
    owner: root
    group: root
    content: |
      #!/bin/bash -x
      echo "Running: $0"
      mkdir -p /usr/local/bin
      if [ ! -x /usr/local/bin/ec2-metadata ]; then
        wget -O /usr/local/bin/ec2-metadata http://s3.amazonaws.com/ec2metadata/ec2-metadata
        chmod ugo+rx /usr/local/bin/ec2-metadata
      fi
      for device in /sys/block/*/queue/rq_affinity ; do echo 2 > $device ; done
      for device in /sys/block/*/queue/scheduler ; do echo noop > $device ; done
      for device in /sys/block/*/queue/nr_requests ; do echo 256 > $device ; done
      for device in /sys/block/*/queue/read_ahead_kb ; do echo 256 > $device ; done
      echo tsc > /sys/devices/system/clocksource/clocksource0/current_clocksource
      # If TLB exhaustion happens on an older core host (newer CPUs shouldn't have this problem)
      #echo never > /sys/kernel/mm/transparent_hugepage/enabled
      sysctl -p
  "/etc/sysctl.d/99_swap.conf":
    mode: "000644"
    owner: root
    group: root
    content: |
      vm.swappiness = 0
  "/etc/sysctl.d/99_filesystem.conf":
    mode: "000644"
    owner: root
    group: root
    content: |
      vm.dirty_ratio = 80
      vm.dirty_background_ratio = 5
      vm.dirty_expire_centisecs = 12000
  "/etc/sysctl.d/99_networking.conf":
    mode: "000644"
    owner: root
    group: root
    content: |
      net.core.somaxconn = 1000
      net.core.netdev_max_backlog = 5000
      net.core.rmem_max = 16777216
      net.core.wmem_max = 16777216
      net.ipv4.tcp_max_syn_backlog = 8096
      net.ipv4.tcp_slow_start_after_idle = 0
      net.ipv4.tcp_tw_reuse = 1
      net.ipv4.ip_local_port_range = 10240 65535
      net.ipv4.tcp_abort_on_overflow = 1