mod_cluster Management Script

Nicolas Gonzalez — Tue, 07 May 2019 14:06:50 +0000

mod_cluster Management Script

mms is a shell script that allows you to execute some of the functions available on Apache's mod_cluster-manager (web).

List all available nodes
Enable, Disable and Stop a Context

Usage

$ chmod +x mms.sh
$ ./mms.sh --help

 ███╗   ███╗███╗   ███╗███████╗
 ████╗ ████║████╗ ████║██╔════╝
 ██╔████╔██║██╔████╔██║███████╗
 ██║╚██╔╝██║██║╚██╔╝██║╚════██║
 ██║ ╚═╝ ██║██║ ╚═╝ ██║███████║
 ╚═╝     ╚═╝╚═╝     ╚═╝╚══════╝

===============================
 MOD_CLUSTER MANAGEMENT SCRIPT
===============================

usage:  ./mms.sh [-h|--host <ip address>] [-p|--port <port>] [-s|--show-nodes]
        ./mms.sh [-h|--host <ip address>] [-p|--port <port>] [-n|--node <node>] { [--context-stop] | [--context-disable] | [--context-enable] } [<Node ID> <Context>]

where:
    (-h|--host)         set the apache mcm host
    (-p|--port)         set the apache mcm port
    (-n|--node)         set the apache mcm node
    (-s|--show-nodes)       list all the nodes
    (--context-enable)      enable a context
    (--context-disable)     disable a context
    (--context-stop)        stop a context

    --help          show this help text

Examples

List all the nodes

    ./mms.sh --host 192.168.4.12 --show-nodes

    Nodes:

    1) wf411 (ajp://192.168.4.11:8009)
      > /TestApp-war        [ENABLED]
      > /   [ENABLED]
      > /wildfly-services   [ENABLED]

    2) wf410 (ajp://192.168.4.10:8009)
      > /TestApp-war        [ENABLED]
      > /   [ENABLED]
      > /wildfly-services   [ENABLED]

Stop a context

    ./mms.sh --host 192.168.4.12 --node 2 --context-stop TestApp-war

    Stopping:

    2) wf410 (ajp://192.168.4.10:8009)
      > /TestApp-war        [ENABLED]

    Disabling from: http://192.168.4.10:5555/mcm?nonce=e87c92bd-195e-4845-91cd-2a192f14a553&Cmd=DISABLE-APP&Range=CONTEXT&JVMRoute=wf410&Alias=default-host&Context=/TestApp-war

    Nodes:

    1) wf411 (ajp://192.168.4.11:8009)
      > /TestApp-war        [ENABLED]
      > /   [ENABLED]
      > /wildfly-services   [ENABLED]

    2) wf410 (ajp://192.168.4.10:8009)
      > /TestApp-war        [STOPPED]
      > /   [ENABLED]
      > /wildfly-services   [ENABLED]

ProGo Bird

Nicolas Gonzalez — Tue, 02 Apr 2019 22:23:55 +0000

https://github.com/nsgonzalez/progo-bird

ProGo Bird is a simple Agent written in Prolog that plans a game for a custom and basic version of Flappy Bird where the bird has to reach a goal. It's pretty basic and it does not calculate what action it should do online, instead it does all the calculation before the game start (calling swipl through python), then store the actions to execute in a slice and pops them until the game is finished.

The game is based on a modified version of Platform under the Pixel's examples (https://github.com/faiface/pixel-examples/tree/master/platformer).

Requisites

debian packages
- git
- golang
- swi-prolog
- python3-pip
- libgl1-mesa-dev
- libxcursor-dev
- libxrandr-dev
- libxinerama-dev
- libxi-dev
pip packages
- pyswip (refactor branch)
OpenGL >= 3.0

Note: it only was tested under Debian 9.4 x86, Ubuntu 18.04.1 x64 and Deepin 15.9.2 x64

Install dependencies

$ sudo apt install -y git golang swi-prolog python3-pip libgl1-mesa-dev libxcursor-dev libxrandr-dev libxinerama-dev libxi-dev
$ wget -O pyswip.zip https://github.com/yuce/pyswip/archive/refactor.zip
$ unzip pyswip.zip
$ cd pyswip-refactor
$ python3 setup.py install --user

How it works

When the game sarts the bird is at (0, 0) and we need it to arrive to a goal located at (X, Y) with a radius R. The speed of the loop in which the game runs is 60fps, so in order to make it work it is multiplied by a factor, which is passed to the agent program so it calculates which moves it should make according to that factor starting from the point (0, 0). Then, the result of the multiplication give us the fps at which the agent performs an action. For example, if the factor would be 0.5, then the result would be 30 and the agent would calculate which actions it should perform every 30fps (0.5 seconds approximately).

I could't find or get successful results with any go library implementing a prolog interpreter, so before the game's loop starts a few calls to a python script implementing pyswip (swipl library for python) is made through a tcp socket and reading back the actions it should perform to play a successful game.

Run

Clone the repository, install the go dependencies and run the program.

$ git clone https://github.com/nsgonzalez/progo-bird.git
$ cd progo-bird
$ go get ./...
$ go run *.go

Successful games

The games that proved to be successful under this weird implementation are the following, and all of them require to modify parameters.go as follows.

Game 1 - 30fps

TIME_AGENT_FACTOR = 0.5
START_AGENT_FACTOR = 4
PLATFORM_MARGIN_H  = 18
PLATFORM_MARGIN_V  = 0
PLATFORM_TB_MARGIN = 0

Game 2 - 24fps

TIME_AGENT_FACTOR = 0.40
START_AGENT_FACTOR = 4
PLATFORM_MARGIN_H  = 10
PLATFORM_MARGIN_V  = 0
PLATFORM_TB_MARGIN = 0

Game 3 - 18fps

TIME_AGENT_FACTOR = 0.30
START_AGENT_FACTOR = 5
PLATFORM_MARGIN_H  = 10
PLATFORM_MARGIN_V  = 0
PLATFORM_TB_MARGIN = 0

Game 4 - 15fps

TIME_AGENT_FACTOR = 0.25
START_AGENT_FACTOR = 4
PLATFORM_MARGIN_H  = 10
PLATFORM_MARGIN_V  = 10
PLATFORM_TB_MARGIN = 10

Game 5 - 12fps

TIME_AGENT_FACTOR = 0.20
START_AGENT_FACTOR = 7
PLATFORM_MARGIN_H  = 0
PLATFORM_MARGIN_V  = 10
PLATFORM_TB_MARGIN = 0

Just a little explanation about the variables (constants really):

TIME_AGENT_FACTOR is the factor that indicates how time is discretized.
START_AGENT_FACTOR is multiplied with TIME_AGENT_FACTOR to simulate the fall when game starts.
PLATFORM_MARGIN_H AND PLATFORM_MARGIN_V are horizontal and vertical margins for the platforms, so the agent doesn't calculate the moves too close to them.
PLATFORM_TB_MARGIN same as above but for the top and bottom platforms.

DEV Community: Nicolas Gonzalez

mod_cluster Management Script

mod_cluster Management Script

Usage

Examples

ProGo Bird

Requisites

Note: it only was tested under Debian 9.4 x86, Ubuntu 18.04.1 x64 and Deepin 15.9.2 x64

Install dependencies

How it works

Run

Successful games

Game 1 - 30fps

Game 2 - 24fps

Game 3 - 18fps

Game 4 - 15fps

Game 5 - 12fps