Send a burst/multi threaded requests using bash and curl - bash

Question

I have a use case where I want to send a burst of requests. My current script sends the request sequentially but I want to be able to send a burst of requests (all of them at once) and wait for the response to come back (ex. if I enter 10 as request count- all 10 should be sent together). I am not sure how I can achieve that using bash and curl. Any ideas would be appreciated.
#!/bin/bash
CURL="/usr/bin/curl"
echo -n "how many times you want to run the request: "
read erc
ERC="$erc"
#count=1;
total_connect=0
total_start=0
total_time=0
echo " Time_Connect Time_startTransfer Time_total HTTP_Code ";
#while [ $count -le $ERC ]
for ((i=1;i<=$ERC;i+=1)); do
result=`$CURL -k -o /dev/null -s -w %{time_connect}:%{time_starttransfer}:%{time_total}:%{http_code} -H "Accept: application/xml" -H "Accept: text/xml" -H "Accept: application/json" -H "Accept: application/cbor" "http://google.com"`
echo $result;
var=$(echo $result | awk -F":" '{print $1, $2, $3, $4}')
set -- $var
total_connect=`echo "scale=6; $total_connect + $1" | bc`;
total_start=`echo "scale=6; $total_start + $2" | bc`;
total_time=`echo "scale=6; $total_time + $3" | bc`;
#count=$((count+1))
done
echo "URL executed is http://google.com"
echo "HTTP CODE of all request is $4"
echo "average time connect: `echo "scale=6; $total_connect/$ERC" | bc`";
echo "average time start: `echo "scale=6; $total_start/$ERC" | bc`";
echo "average Totaltime taken: `echo "scale=6; $total_time/$ERC" | bc`";

Answer 1

Here's the code with the & introduced to issue all of the curls "at the same time".
#!/bin/bash
CURL="/usr/bin/curl"
echo -n "how many times you want to run the request: "
read erc
ERC="$erc"
#count=1;
total_connect=0
total_start=0
total_time=0
echo " Time_Connect Time_startTransfer Time_total HTTP_Code ";
#while [ $count -le $ERC ]
for ((i=1;i<=$ERC;i+=1)); do
{
result=`$CURL -k -o /dev/null -s -w %{time_connect}:%{time_starttransfer}:%{time_total}:%{http_code} -H "Accept: application/xml" -H "Accept: text/xml" -H "Accept: application/json" -H "Accept: application/cbor" "http://google.com"`
echo $result;
var=$(echo $result | awk -F":" '{print $1, $2, $3, $4}')
set -- $var
total_connect=`echo "scale=6; $total_connect + $1" | bc`;
total_start=`echo "scale=6; $total_start + $2" | bc`;
total_time=`echo "scale=6; $total_time + $3" | bc`;
#count=$((count+1))
} > $i.result &
done
wait
# TODO: You'll need to write code here to process the intermediate results in *.result
# It would have been helpful to have created a temporary directory for your indivdual results as well
echo "URL executed is http://google.com"
echo "HTTP CODE of all request is $4"
echo "average time connect: `echo "scale=6; $total_connect/$ERC" | bc`";
echo "average time start: `echo "scale=6; $total_start/$ERC" | bc`";
echo "average Totaltime taken: `echo "scale=6; $total_time/$ERC" | bc`";
The wait I introduced tells bash to wait for all of the child processes (curls and such) to complete before continuing.