Syslog support

[vvviperrr]

syslog logging is useful in embedded applications, e.g in buildroot software

[rfjakob]

Hi, thanks for the PR!

But first I have to ask a stupid question: Why is this better than piping stdout and stderr to logger(1)?

[vvviperrr]

Hi, thanks for the PR!

But first I have to ask a stupid question: Why is this better than piping stdout and stderr to logger(1)?

Hi. The main reason - support different syslog priorities. Sure, we can achieve this via piping earlyoom's stdout & stderr in different logger instances with different facilities\levels, but this trick is really ugly. something like this (example from busybox sysctl script)

    xargs -0 -r -n 1 readlink -f | {
        prog_status="OK"
        while :; do
            read -r file || {
                echo "$prog_status" >&4
                break
            }
            echo "* Applying $file ..."
            /sbin/sysctl -p "$file" $SYSCTL_ARGS || prog_status="FAIL"
        done 2>&1 >&3 | /usr/bin/logger -t sysctl -p kern.err
    } 3>&1 | /usr/bin/logger -t sysctl -p kern.info

the things become more complicated, when u need start earlyoom via start-stop-daemon.

so patching earlyoom code is just handy (by adding syslog param in default conf file) & portable solution.

[rfjakob]

Ok, understood. I will accept syslog support.

[rfjakob]

I merged this, thanks. Two more things if you have time:

1) Can you add --syslog to https://github.com/rfjakob/earlyoom/blob/master/MANPAGE.md

2) The startup messages still go to stdout, seems to be they should be going to syslog as well:

$ ./earlyoom --syslog
earlyoom v1.7-36-g848653f
mem total: 23881 MiB, user mem total: 21593 MiB, swap total: 8191 MiB
sending SIGTERM when mem <= 10.00% and swap <= 10.00%,
        SIGKILL when mem <=  5.00% and swap <=  5.00%