This post is more than 5 years old
5 Practitioner
•
274.2K Posts
0
1505
syslogd troubleshooting
So I am having this issue with the syslogd daemon. I had to restart the syslogd daemon. It did not work whenever I did a isi services -a syslogd disable. Therefore I tried another way that was suggested to me which was the killall syslogd. For some reason whenever I ran the isi_for_array killall syslogd command, the daemon started to run with a -L option. I have no idea what that option means, it is also not anywhere in the man pages. Please help!!
kipcranford
125 Posts
0
March 23rd, 2017 15:00
Hi there.
Not sure which release you're running, but "recently" (I think around the OneFS 7.2 days) we've added the -L option to syslogd. When present, this option controls the addition of the ID suffix to a node hostname, i.e. it adds the node ID (which is different than the node LNN, and is unique for the life of the cluster) such that messages aggregated on an external server can be traced back to their originating node.
When you killed syslogd manually, MCP restarted it for you (that's the facility that monitors are well-known services) and added that option. You can see this from the MCP configuration:
isilon# cat /etc/mcp/sys/services/syslogd
System Logging Daemon
startaction="start" stopaction="stop"/>
/bin/rm -f /var/run/log
/usr/sbin/syslogd -c -L
/bin/kill `cat /var/run/syslog.pid`
(note, don't play around with files in /etc/mcp/sys/, or really in any /etc/mcp/ directory imo, without at least consulting Isilon Support).