tpotce/bin/restore_es.sh

#/bin/bash
# Restore folder based ES backup
# Make sure ES is available
myES="http://127.0.0.1:64298/"
myESSTATUS=$(curl -s -XGET ''$myES'_cluster/health' | jq '.' | grep -c green)
if ! [ "$myESSTATUS" = "1" ]
  then
    echo "### Elasticsearch is not available, try starting via 'systemctl start elk'."
    exit
  else
    echo "### Elasticsearch is available, now continuing."
fi

# Let's ensure normal operation on exit or if interrupted ...
function fuCLEANUP {
  rm -rf tmp
}
trap fuCLEANUP EXIT

# Set vars
myDUMP=$1
myCOL1=""
myCOL0=""

# Check if parameter is given and file exists
if [ "$myDUMP" = "" ];
    then
      echo $myCOL1"### Please provide a backup file name."$myCOL0 
      echo $myCOL1"### restore-elk.sh <es_dump.tar>"$myCOL0
      echo 
      exit 
fi
if ! [ -a $myDUMP ];
    then
      echo $myCOL1"### File not found."$myCOL0 
      exit
fi

# Unpack tar archive
echo $myCOL1"### Now unpacking tar archive: "$myDUMP $myCOL0
tar xvf $myDUMP

# Build indices list
myINDICES=$(ls tmp/logstash*.gz | cut -c 5- | rev | cut -c 4- | rev)
myINDICES+=" .kibana"
echo $myCOL1"### The following indices will be restored: "$myCOL0
echo $myINDICES
echo

# Restore indices
for i in $myINDICES;
  do
    # Delete index if it already exists
    curl -s -XDELETE $myES$i > /dev/null
    echo $myCOL1"### Now uncompressing: tmp/$i.gz" $myCOL0
    gunzip -f tmp/$i.gz
    # Restore index to ES
    echo $myCOL1"### Now restoring: "$i $myCOL0
    elasticdump --input=tmp/$i --output=$myES$i --limit 7500
    rm tmp/$i
  done;
echo $myCOL1"### Done."$myCOL0