I am using gradle for build and release, so my gradle script executes a shell script. The shell script outputs an ip address which has to be provided as an input to my next gradle ssh task. I am able to get the output and print on the console but not able to use this output as an input to next task.
remotes {
web01 {
def ip = exec {
commandLine './returnid.sh'
}
println ip --> i am able to see the ip address on console
role 'webServers'
host = ip --> i tried referring as $ip '$ip' , both results into syntax error
user = 'ubuntu'
password = 'ubuntu'
}
}
task checkWebServers1 << {
ssh.run {
session(remotes.web01) {
execute 'mkdir -p /home/ubuntu/abc3'
}
}
}
but it results in error "
What went wrong:
Execution failed for task ':checkWebServers1'.
java.net.UnknownHostException: {exitValue=0, failure=null}"
Can anyone please help me use the output variable in proper syntax or provide some hints which could help me.
Thanks in advance
The reason it's not working is the fact, that exec
call return is ExecResult
(here is it's JavaDoc description) and it's not a text output of the execution.
If you need to get the text output, then you've to specify the standardOutput
property of the exec
task. This could be done so:
remotes {
web01 {
def ip = new ByteArrayOutputStream()
exec {
commandLine './returnid.sh'
standardOutput = ip
}
println ip
role 'webServers'
host = ip.toString().split("\n")[2].trim()
user = 'ubuntu'
password = 'ubuntu'
}
}
Just note, the ip value by default would have a multiline output, include the command itself, so it has to be parsed to get the correct output, For my Win machine, this could be done as:
ip.toString().split("\n")[2].trim()
Here it takes only first line of the output.