Spong client is the client application that works with SpongServer which we use to monitor our servers and nodes.
Here are some installation instructions for getting it working on your node, they are specific to DebianLinux but should be fairly easy to adapt. Note that spong-client doesn't currently install on a Red Hat Linux machine from the tarball.
Set the host name of your box to your NodeId.
# hostname node111 BR # vi /etc/hostname The /etc/hostname file should include one line {{{ node111
}}}
- Install spong client onto your box.
apt-get install spong-client
Set your spong server to mon.personaltelco.net.
- Setup your domain name to be personaltelco.net and make sure that a gethostbyname call returns the fully qualified domain name (eg. node111.personaltelco.net).
# vi /etc/hosts It should be like this (if you have other entries that's okay but make sure that these two exist and are exactly like this): {{{127.0.0.1 localhost x.x.x.x node111.personaltelco.net node111}}}
- You can test it to make sure it's doing the right thing by running this perl one liner:
# perl -MSys::Hostname -e '$h = (gethostbyname(Sys::Hostname::hostname()))[0]; print "$h\n"'
- Now setup spong to monitor what you want it to. The main thing you need to change is what processes it makes sure are running.
# vi /etc/spong/spong.conf Look for the line which starts @PROCS and add the processes that you want it look for. Here's a starting point: {{{@PROCSWARN = ("cron","atd");
- @PROCSCRIT = ("syslogd","klogd","dhcpd","cardmgr","gateway")}}}
- Now test to make sure it's all working okay (errors should be obvious).
# /usr/sbin/spong-client --debug
Email mailto:ops@personaltelco.net all this information so we can add your system to be monitored. We need your NodeId, your IpAddress, your name and whatever email address you'd like to be notified of failures on.
- Check the spong page and make sure that you are now listed.