fabric hangs up some remote command (for daemon program)

[bitprophet]

Description

When I use fabric deploy configuration files and restart services for Ganglia, I found some ganglia monitor processes hanged up.

deploy script:

41 def refresh_DEV_node():
42     put('/home/hadoop/workspace/ganglia/conf/DEV_node/*', '/etc/ganglia', True)
43     sudo('/etc/init.d/ganglia-monitor restart')

fabric version:

Fabric 1.0.1

Ganglia version:

gmond 3.1.7

Linux box infomation:

Linux Debian 2.6.32-5-amd64 #1 SMP Wed Jan 12 03:40:32 UTC 2011 x86_64 GNU/Linux

---

Originally submitted by **** (docete) on 2011-05-11 at 10:21pm EDT

Relations

• Related to Init scripts frequently fail to start their daemons #395: Init scripts frequently fail to start their daemons

[bitprophet]

Max Arnold (LwarX) posted:

---

Zavier Gao: have you tried to pass pty=False to the sudo() call? I have the following function:

def restart_apache():
    sudo('/etc/init.d/apache2 restart')

[web1.internal] Executing task 'restart_apache'
[web1.internal] sudo: /etc/init.d/apache2 restart
[web1.internal] out:  * Restarting web server apache2
[web1.internal] out:    ...done.
[web1.internal] out: 
Done.
Disconnecting from web1.internal... done.

Output looks normal, but actually apache haven't started. Adding pty=False helps. Remote OS is Ubuntu Lucid.

---

on 2011-05-15 at 06:31am EDT

[bitprophet]

**** (docete) posted:

---

Thanks Arnold. have tried but didn't work for ganglia.
Btw:
my situation is:　I have many machines(tens of with the same hardware/software), the command works for some of them but doesn't work for the others.

---

on 2011-05-19 at 03:09am EDT

[bitprophet]

Considering this a specific case of #395.