#!/bin/bash
: '
 @vendor    Chorke Academia, Inc.
 @web       https://cdn.chorke.org/docs/academia
 @version   1.0.00
 @since     1.0.00
'

:<<EOF
if [ -x "$(command -v curl)" ];then \
export HADOOP_HOME_DIR='/var/hadoop/hadoop-3.3.4';\
export MINMAX_CAPACITIES='([dev]="(50 50)" [prod]="(50 70)")';\
export HADOOP_SCHEDULERS='([dev]="Development" [prod]="Production")';\
bash <(curl -s 'https://cdn.chorke.org/exec/cli/bash/install/apache-hadoop/3.3.4-scheduler.sh.txt');\
else printf 'curl \033[0;31mnot found! \033[0m:(\n';fi
EOF


function conf_yarn_site(){
    # exported varriable to indexed & associative array
    declare -A HADOOP_SCHEDULERS_MAP="${HADOOP_SCHEDULERS}"
    declare -A MINMAX_CAPACITIES_MAP="${MINMAX_CAPACITIES}"
    declare -a HADOOP_SCHEDULERS_SET="($(echo ${!HADOOP_SCHEDULERS_MAP[@]}|tr ' ' $'\n'|sort))"


    # add yarn config to yarn-site.xml
    YARN_SITE_FILE="${HADOOP_HOME_DIR}/etc/hadoop/yarn-site.xml"

    YARN_CONF_FIND=$(cat <<EOF
<configuration>\n\n\
<!-- Site specific YARN configuration properties -->\n\n\
</configuration>
EOF
)
    YARN_CONF_INDX=0
    declare -a YARN_CONF_FILL
    YARN_CONF_FILL[${YARN_CONF_INDX}]=$(cat <<EOF
<configuration>\n\
  <property>\n\
    <name>yarn.resourcemanager.scheduler.class</name>\n\
    <value>org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.CapacityScheduler</value>\n\
  </property>\n\
  <property>\n\
    <name>yarn.scheduler.capacity.root.queues</name>\n\
    <value>$(echo "${HADOOP_SCHEDULERS_SET[@]}"|tr ' ' ',')</value>\n\
  </property>
EOF
)
    for HADOOP_SCHEDULER in ${HADOOP_SCHEDULERS_SET[@]};do
        declare -a MINMAX_CAPACITY="${MINMAX_CAPACITIES_MAP[${HADOOP_SCHEDULER}]}"
        MIN_CAPACITY=${MINMAX_CAPACITY[0]}

        YARN_CONF_INDX=$((${YARN_CONF_INDX}+1))
        YARN_CONF_FILL[${YARN_CONF_INDX}]=$(cat <<EOF
\n\
  <property>\n\
    <name>yarn.scheduler.capacity.${HADOOP_SCHEDULER}.capacity</name>\n\
    <value>${MIN_CAPACITY}</value>\n\
  </property>
EOF
)
    done
    for HADOOP_SCHEDULER in ${HADOOP_SCHEDULERS_SET[@]};do
        declare -a MINMAX_CAPACITY="${MINMAX_CAPACITIES_MAP[${HADOOP_SCHEDULER}]}"
        MAX_CAPACITY=${MINMAX_CAPACITY[1]}

        YARN_CONF_INDX=$((${YARN_CONF_INDX}+1))
        YARN_CONF_FILL[${YARN_CONF_INDX}]=$(cat <<EOF
\n\
  <property>\n\
    <name>yarn.scheduler.capacity.${HADOOP_SCHEDULER}.maximum-capacity</name>\n\
    <value>${MAX_CAPACITY}</value>\n\
  </property>
EOF
)
    done
    YARN_CONF_INDX=$((${YARN_CONF_INDX}+1))
    YARN_CONF_FILL[${YARN_CONF_INDX}]=$(cat <<EOF
\n\
</configuration>
EOF
)
    sudo sed -z "s|${YARN_CONF_FIND}|$(echo "${YARN_CONF_FILL[@]}")|" -i "${YARN_SITE_FILE}"
}


function conf_capacity_scheduler(){
    # exported varriable to indexed & associative array
    declare -A HADOOP_SCHEDULERS_MAP="${HADOOP_SCHEDULERS}"
    declare -A MINMAX_CAPACITIES_MAP="${MINMAX_CAPACITIES}"
    declare -a HADOOP_SCHEDULERS_SET="($(echo ${!HADOOP_SCHEDULERS_MAP[@]}|tr ' ' $'\n'|sort))"


    # add capacity config to capacity-scheduler.xml
    CAPACITY_SCHEDULER_FILE="${HADOOP_HOME_DIR}/etc/hadoop/capacity-scheduler.xml"


    # replace default root queues
    ROOT_QUEUES_FIND=$(cat <<EOF
[ ]*<name>yarn.scheduler.capacity.root.queues</name>\n\
[ ]*<value>default</value>
EOF
)
    ROOT_QUEUES_FILL=$(cat <<EOF
    <name>yarn.scheduler.capacity.root.queues</name>\n\
    <value>$(echo "${HADOOP_SCHEDULERS_SET[@]}"|tr ' ' ',')</value>
EOF
)
    sed -z "s|${ROOT_QUEUES_FIND}|${ROOT_QUEUES_FILL}|" -i "${CAPACITY_SCHEDULER_FILE}"


    # replace default root capacity
    DEFAULT_CAPACITY_FIND=$(cat <<EOF
\n\
  <property>\n\
    <name>yarn.scheduler.capacity.root.default.capacity</name>\n\
    <value>100</value>\n\
    <description>Default queue target capacity.</description>\n\
  </property>
EOF
)
    DEFAULT_CAPACITY_INDX=0
    declare -a DEFAULT_CAPACITY_FILL
    for HADOOP_SCHEDULER in ${HADOOP_SCHEDULERS_SET[@]};do
        declare -a MINMAX_CAPACITY="${MINMAX_CAPACITIES_MAP[${HADOOP_SCHEDULER}]}"
        HADOOP_SCHEDULER_NAME=${HADOOP_SCHEDULERS_MAP[${HADOOP_SCHEDULER}]}
        MIN_CAPACITY=${MINMAX_CAPACITY[0]}

        DEFAULT_CAPACITY_FILL[${DEFAULT_CAPACITY_INDX}]=$(cat <<EOF
\n\
  <property>\n\
    <name>yarn.scheduler.capacity.root.${HADOOP_SCHEDULER}.capacity</name>\n\
    <value>${MIN_CAPACITY}</value>\n\
    <description>${HADOOP_SCHEDULER_NAME} queue target capacity.</description>\n\
  </property>
EOF
)
        DEFAULT_CAPACITY_INDX=$((${DEFAULT_CAPACITY_INDX}+1))
    done
    sed -z "s|${DEFAULT_CAPACITY_FIND}|$(echo "${DEFAULT_CAPACITY_FILL[@]}")|" -i "${CAPACITY_SCHEDULER_FILE}"


    # replace default root user limit factor
    DEFAULT_USER_LIMIT_FACTOR_FIND=$(cat <<EOF
\n\
  <property>\n\
    <name>yarn.scheduler.capacity.root.default.user-limit-factor</name>\n\
    <value>1</value>\n\
    <description>\n\
      Default queue user limit a percentage from 0.0 to 1.0.\n\
    </description>\n\
  </property>
EOF
)
    DEFAULT_USER_LIMIT_FACTOR_INDX=0
    declare -a DEFAULT_USER_LIMIT_FACTOR_FILL
    for HADOOP_SCHEDULER in ${HADOOP_SCHEDULERS_SET[@]};do
        HADOOP_SCHEDULER_NAME=${HADOOP_SCHEDULERS_MAP[${HADOOP_SCHEDULER}]}
        DEFAULT_USER_LIMIT_FACTOR_FILL[${DEFAULT_USER_LIMIT_FACTOR_INDX}]=$(cat <<EOF
\n\
  <property>\n\
    <name>yarn.scheduler.capacity.root.${HADOOP_SCHEDULER}.user-limit-factor</name>\n\
    <value>1</value>\n\
    <description>\n\
      ${HADOOP_SCHEDULER_NAME} queue user limit a percentage from 0.0 to 1.0.\n\
    </description>\n\
  </property>
EOF
)
        DEFAULT_USER_LIMIT_FACTOR_INDX=$((${DEFAULT_USER_LIMIT_FACTOR_INDX}+1))
    done
    sed -z "s|${DEFAULT_USER_LIMIT_FACTOR_FIND}|$(echo "${DEFAULT_USER_LIMIT_FACTOR_FILL[@]}")|" -i "${CAPACITY_SCHEDULER_FILE}"


    # replace default root maximum capacity
    DEFAULT_MAX_CAPACITY_FIND=$(cat <<EOF
\n\
  <property>\n\
    <name>yarn.scheduler.capacity.root.default.maximum-capacity</name>\n\
    <value>100</value>\n\
    <description>\n\
      The maximum capacity of the default queue. \n\
    </description>\n\
  </property>
EOF
)
    DEFAULT_MAX_CAPACITY_INDX=0
    declare -a DEFAULT_MAX_CAPACITY_FILL
    for HADOOP_SCHEDULER in ${HADOOP_SCHEDULERS_SET[@]};do
        declare -a MINMAX_CAPACITY="${MINMAX_CAPACITIES_MAP[${HADOOP_SCHEDULER}]}"
        HADOOP_SCHEDULER_NAME=${HADOOP_SCHEDULERS_MAP[${HADOOP_SCHEDULER}]}
        MAX_CAPACITY=${MINMAX_CAPACITY[1]}

        DEFAULT_MAX_CAPACITY_FILL[${DEFAULT_MAX_CAPACITY_INDX}]=$(cat <<EOF
\n\
  <property>\n\
    <name>yarn.scheduler.capacity.root.${HADOOP_SCHEDULER}.maximum-capacity</name>\n\
    <value>${MAX_CAPACITY}</value>\n\
    <description>The maximum capacity of the ${HADOOP_SCHEDULER_NAME} queue.</description>\n\
  </property>
EOF
)
        DEFAULT_MAX_CAPACITY_INDX=$((${DEFAULT_MAX_CAPACITY_INDX}+1))
    done
    sed -z "s|${DEFAULT_MAX_CAPACITY_FIND}|$(echo "${DEFAULT_MAX_CAPACITY_FILL[@]}")|" -i "${CAPACITY_SCHEDULER_FILE}"


    # replace default root state
    DEFAULT_STATE_FIND=$(cat <<EOF
\n\
  <property>\n\
    <name>yarn.scheduler.capacity.root.default.state</name>\n\
    <value>RUNNING</value>\n\
    <description>\n\
      The state of the default queue. State can be one of RUNNING or STOPPED.\n\
    </description>\n\
  </property>
EOF
)
    DEFAULT_STATE_INDX=0
    declare -a DEFAULT_STATE_FILL
    for HADOOP_SCHEDULER in ${HADOOP_SCHEDULERS_SET[@]};do
        HADOOP_SCHEDULER_NAME=${HADOOP_SCHEDULERS_MAP[${HADOOP_SCHEDULER}]}
        DEFAULT_STATE_FILL[${DEFAULT_STATE_INDX}]=$(cat <<EOF
\n\
  <property>\n\
    <name>yarn.scheduler.capacity.root.${HADOOP_SCHEDULER}.state</name>\n\
    <value>RUNNING</value>\n\
    <description>\n\
      The state of the ${HADOOP_SCHEDULER_NAME} queue. State can be one of RUNNING or STOPPED.\n\
    </description>\n\
  </property>
EOF
)
        DEFAULT_STATE_INDX=$((${DEFAULT_STATE_INDX}+1))
    done
    sed -z "s|${DEFAULT_STATE_FIND}|$(echo "${DEFAULT_STATE_FILL[@]}")|" -i "${CAPACITY_SCHEDULER_FILE}"


    # replace default root acl submit applications
    DEFAULT_ACL_SUBMIT_APPLICATIONS_FIND=$(cat <<EOF
\n\
  <property>\n\
    <name>yarn.scheduler.capacity.root.default.acl_submit_applications</name>\n\
    <value>\*</value>\n\
    <description>\n\
      The ACL of who can submit jobs to the default queue.\n\
    </description>\n\
  </property>
EOF
)
    DEFAULT_ACL_SUBMIT_APPLICATIONS_INDX=0
    declare -a DEFAULT_ACL_SUBMIT_APPLICATIONS_FILL
    for HADOOP_SCHEDULER in ${HADOOP_SCHEDULERS_SET[@]};do
        HADOOP_SCHEDULER_NAME=${HADOOP_SCHEDULERS_MAP[${HADOOP_SCHEDULER}]}
        DEFAULT_ACL_SUBMIT_APPLICATIONS_FILL[${DEFAULT_ACL_SUBMIT_APPLICATIONS_INDX}]=$(cat <<EOF
\n\
  <property>\n\
    <name>yarn.scheduler.capacity.root.${HADOOP_SCHEDULER}.acl_submit_applications</name>\n\
    <value>*</value>\n\
    <description>\n\
      The ACL of who can submit jobs to the ${HADOOP_SCHEDULER_NAME} queue.\n\
    </description>\n\
  </property>
EOF
)
        DEFAULT_ACL_SUBMIT_APPLICATIONS_INDX=$((${DEFAULT_ACL_SUBMIT_APPLICATIONS_INDX}+1))
    done
    sed -z "s|${DEFAULT_ACL_SUBMIT_APPLICATIONS_FIND}|$(echo "${DEFAULT_ACL_SUBMIT_APPLICATIONS_FILL[@]}")|" -i "${CAPACITY_SCHEDULER_FILE}"


    # replace default root acl administer queue
    DEFAULT_ACL_ADMINISTER_QUEUE_FIND=$(cat <<EOF
\n\
  <property>\n\
    <name>yarn.scheduler.capacity.root.default.acl_administer_queue</name>\n\
    <value>\*</value>\n\
    <description>\n\
      The ACL of who can administer jobs on the default queue.\n\
    </description>\n\
  </property>
EOF
)
    DEFAULT_ACL_ADMINISTER_QUEUE_INDX=0
    declare -a DEFAULT_ACL_ADMINISTER_QUEUE_FILL
    for HADOOP_SCHEDULER in ${HADOOP_SCHEDULERS_SET[@]};do
        HADOOP_SCHEDULER_NAME=${HADOOP_SCHEDULERS_MAP[${HADOOP_SCHEDULER}]}
        DEFAULT_ACL_ADMINISTER_QUEUE_FILL[${DEFAULT_ACL_ADMINISTER_QUEUE_INDX}]=$(cat <<EOF
\n\
  <property>\n\
    <name>yarn.scheduler.capacity.root.${HADOOP_SCHEDULER}.acl_administer_queue</name>\n\
    <value>*</value>\n\
    <description>\n\
      The ACL of who can administer jobs on the ${HADOOP_SCHEDULER_NAME} queue.\n\
    </description>\n\
  </property>
EOF
)
        DEFAULT_ACL_ADMINISTER_QUEUE_INDX=$((${DEFAULT_ACL_ADMINISTER_QUEUE_INDX}+1))
    done
    sed -z "s|${DEFAULT_ACL_ADMINISTER_QUEUE_FIND}|$(echo "${DEFAULT_ACL_ADMINISTER_QUEUE_FILL[@]}")|" -i "${CAPACITY_SCHEDULER_FILE}"


    # replace default root acl application max priority
    DEFAULT_ACL_APPLICATION_MAX_PRIORITY_FIND=$(cat <<EOF
\n\
  <property>\n\
    <name>yarn.scheduler.capacity.root.default.acl_application_max_priority</name>\n\
    <value>\*</value>\n\
    <description>\n\
      The ACL of who can submit applications with configured priority.\n\
      For e.g, \[user={name} group={name} max_priority={priority} default_priority={priority}\]\n\
    </description>\n\
  </property>
EOF
)
    DEFAULT_ACL_APPLICATION_MAX_PRIORITY_INDX=0
    declare -a DEFAULT_ACL_APPLICATION_MAX_PRIORITY_FILL
    for HADOOP_SCHEDULER in ${HADOOP_SCHEDULERS_SET[@]};do
        DEFAULT_ACL_APPLICATION_MAX_PRIORITY_FILL[${DEFAULT_ACL_APPLICATION_MAX_PRIORITY_INDX}]=$(cat <<EOF
\n\
  <property>\n\
    <name>yarn.scheduler.capacity.root.${HADOOP_SCHEDULER}.acl_application_max_priority</name>\n\
    <value>*</value>\n\
    <description>\n\
      The ACL of who can submit applications with configured priority.\n\
      For e.g, [user={name} group={name} max_priority={priority} default_priority={priority}]\n\
    </description>\n\
  </property>
EOF
)
        DEFAULT_ACL_APPLICATION_MAX_PRIORITY_INDX=$((${DEFAULT_ACL_APPLICATION_MAX_PRIORITY_INDX}+1))
    done
    sed -z "s|${DEFAULT_ACL_APPLICATION_MAX_PRIORITY_FIND}|$(echo "${DEFAULT_ACL_APPLICATION_MAX_PRIORITY_FILL[@]}")|" -i "${CAPACITY_SCHEDULER_FILE}"


    # replace default root maximum application lifetime
    DEFAULT_MAXIMUM_APPLICATION_LIFETIME_FIND=$(cat <<EOF
\n\
   <property>\n\
     <name>yarn.scheduler.capacity.root.default.maximum-application-lifetime\n\
     </name>\n\
     <value>-1</value>\n\
     <description>\n\
        Maximum lifetime of an application which is submitted to a queue\n\
        in seconds. Any value less than or equal to zero will be considered as\n\
        disabled.\n\
        This will be a hard time limit for all applications in this\n\
        queue. If positive value is configured then any application submitted\n\
        to this queue will be killed after exceeds the configured lifetime.\n\
        User can also specify lifetime per application basis in\n\
        application submission context. But user lifetime will be\n\
        overridden if it exceeds queue maximum lifetime. It is point-in-time\n\
        configuration.\n\
        Note : Configuring too low value will result in killing application\n\
        sooner. This feature is applicable only for leaf queue.\n\
     </description>\n\
   </property>
EOF
)
    DEFAULT_MAXIMUM_APPLICATION_LIFETIME_INDX=0
    declare -a DEFAULT_MAXIMUM_APPLICATION_LIFETIME_FILL
    for HADOOP_SCHEDULER in ${HADOOP_SCHEDULERS_SET[@]};do
        DEFAULT_MAXIMUM_APPLICATION_LIFETIME_FILL[${DEFAULT_MAXIMUM_APPLICATION_LIFETIME_INDX}]=$(cat <<EOF
\n\
   <property>\n\
     <name>yarn.scheduler.capacity.root.${HADOOP_SCHEDULER}.maximum-application-lifetime\n\
     </name>\n\
     <value>-1</value>\n\
     <description>\n\
        Maximum lifetime of an application which is submitted to a queue\n\
        in seconds. Any value less than or equal to zero will be considered as\n\
        disabled.\n\
        This will be a hard time limit for all applications in this\n\
        queue. If positive value is configured then any application submitted\n\
        to this queue will be killed after exceeds the configured lifetime.\n\
        User can also specify lifetime per application basis in\n\
        application submission context. But user lifetime will be\n\
        overridden if it exceeds queue maximum lifetime. It is point-in-time\n\
        configuration.\n\
        Note : Configuring too low value will result in killing application\n\
        sooner. This feature is applicable only for leaf queue.\n\
     </description>\n\
   </property>
EOF
)
        DEFAULT_MAXIMUM_APPLICATION_LIFETIME_INDX=$((${DEFAULT_MAXIMUM_APPLICATION_LIFETIME_INDX}+1))
    done
    sed -z "s|${DEFAULT_MAXIMUM_APPLICATION_LIFETIME_FIND}|$(echo "${DEFAULT_MAXIMUM_APPLICATION_LIFETIME_FILL[@]}")|" -i "${CAPACITY_SCHEDULER_FILE}"


    # replace default root default application lifetime
    DEFAULT_APPLICATION_LIFETIME_FIND=$(cat <<EOF
\n\
   <property>\n\
     <name>yarn.scheduler.capacity.root.default.default-application-lifetime\n\
     </name>\n\
     <value>-1</value>\n\
     <description>\n\
        Default lifetime of an application which is submitted to a queue\n\
        in seconds. Any value less than or equal to zero will be considered as\n\
        disabled.\n\
        If the user has not submitted application with lifetime value then this\n\
        value will be taken. It is point-in-time configuration.\n\
        Note : Default lifetime can't exceed maximum lifetime. This feature is\n\
        applicable only for leaf queue.\n\
     </description>\n\
   </property>
EOF
)
    DEFAULT_APPLICATION_LIFETIME_INDX=0
    declare -a DEFAULT_APPLICATION_LIFETIME_FILL
    for HADOOP_SCHEDULER in ${HADOOP_SCHEDULERS_SET[@]};do
        DEFAULT_APPLICATION_LIFETIME_FILL[${DEFAULT_APPLICATION_LIFETIME_INDX}]=$(cat <<EOF
\n\
   <property>\n\
     <name>yarn.scheduler.capacity.root.${HADOOP_SCHEDULER}.default-application-lifetime\n\
     </name>\n\
     <value>-1</value>\n\
     <description>\n\
        Default lifetime of an application which is submitted to a queue\n\
        in seconds. Any value less than or equal to zero will be considered as\n\
        disabled.\n\
        If the user has not submitted application with lifetime value then this\n\
        value will be taken. It is point-in-time configuration.\n\
        Note : Default lifetime can't exceed maximum lifetime. This feature is\n\
        applicable only for leaf queue.\n\
     </description>\n\
   </property>
EOF
)
        DEFAULT_APPLICATION_LIFETIME_INDX=$((${DEFAULT_APPLICATION_LIFETIME_INDX}+1))
    done
    sed -z "s|${DEFAULT_APPLICATION_LIFETIME_FIND}|$(echo "${DEFAULT_APPLICATION_LIFETIME_FILL[@]}")|" -i "${CAPACITY_SCHEDULER_FILE}"
}


function conf_init(){
    conf_yarn_site
    conf_capacity_scheduler
}


function void_init(){
    if [[ "${HADOOP_HOME_DIR}/env/set" == '/env/set' ]];then
        export HADOOP_HOME_DIR='/var/hadoop/hadoop-3.3.4'
    fi

    if [[ "${MINMAX_CAPACITIES}/env/set" == '/env/set' ]];then
        export MINMAX_CAPACITIES="([dev]='(50 50)' [prod]='(50 70)')"
    fi

    if [[ "${HADOOP_SCHEDULERS}/env/set" == '/env/set' ]];then
        export HADOOP_SCHEDULERS='([dev]="Development" [prod]="Production")'
    fi
}


function init(){
    void_init
    conf_init
}


init