Sometimes "kill job" cannot kill the process actually

[ArchestMage]

Hi,
I have a job configured to execute a shell script for maven compiling.
Sometimes,when the job is running and  maven is just downloading some
jars , I kill the job. And through the web UI,I can see that I have kill the job
successfully. But on the target host I see that the process is still there.
Why?
Thank you very much.


ArchestMage

[Moses Lei]

Hi ArchestMage, this is because the default configuration of RunDeck does not ask for a pseudo-tty when logging in to the target box via SSH, thus there isn't a way to propagate a signal to the target node.

If this is important to you, you can configure the script-based node executor to use OpenSSH instead, and use "ssh -tt" to force allocation of a pseudo-tty. Then killing the job should kill the ssh process and also the running processes on the client that belong to that tty.

Moses

--

Moses Lei

Principal, Village Chime LLC

mobile: +1 703 901 5969 | skype: moseslei | yahoo: moseslei

ArchestMage

--
You received this message because you are subscribed to the Google Groups "rundeck-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rundeck-discu...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Moses Lei]

Note that killing the job will work on the RunDeck server, just not on remote clients.

--

Moses Lei

Principal, Village Chime LLC

mobile: +1 703 901 5969 | skype: moseslei | yahoo: moseslei

[Daniel Valente]

Hello,

  I've configured Rundeck with script-based node executor to use OpenSSH, but kill still don't work when I dispatch remotly a command.

  I use: (plugin.script-exec.default.command=ssh -q -tt ${node.username}@${node.hostname} ${exec.command})
  Is there any other option to enforce kill?

Best Regards