Error "logger: invalid option ... multi-call binary" in network service

[niusmallnan]

RancherOS Version: (ros os version)
v1.5.1

Where are you running RancherOS? (docker-machine, AWS, GCE, baremetal, etc.)
AWS

Boot a VM on AWS, we can get these errors by system-docker logs network:

[root@ip-172-31-14-228 rancher]# system-docker logs network
[            ] netconf:info: Apply Network Config
[            ] netconf:info: Running DHCP on eth0: dhcpcd -MA4 --timeout 10 -w --debug eth0
eth0: MTU set to 9001
logger: invalid option -- 'i'
BusyBox v1.27.2 () multi-call binary.

Usage: logger [OPTIONS] [MESSAGE]

Write MESSAGE (or stdin) to syslog

	-s	Log to stderr as well as the system log
	-t TAG	Log using the specified tag (defaults to user name)
	-p PRIO	Priority (numeric or facility.level pair)
[            ] netconf:info: Checking to see if DNS was set by DHCP
[            ] netconf:info: dns testing eth0
[            ] netconf:info: dns was dhcp set for eth0
[            ] netconf:info: DNS set by DHCP
[            ] netconf:info: Apply Network Config SyncHostname
[            ] netconf:info: Restart syslog

This is caused by 10-mtu(/lib/dhcpcd/dhcpcd-hooks/10-mtu) in dhcpcd-hook.

Now dhcpcd can handle the mtu option without this hook, so I will disable this mtu hook by default.

[niusmallnan]

If I remove the /lib/dhcpcd/dhcpcd-hooks/10-mtu, the mtu can be applied to routes.

[root@ip-172-31-8-172 dhcpcd-hooks]# ip route
default via 172.31.0.1 dev eth0 src 172.31.8.172 metric 202 mtu 9001
172.17.0.0/16 dev docker0 proto kernel scope link src 172.17.0.1 linkdown

But I still get 1500 from ip link show:

[root@ip-172-31-14-228 boot]# ip link show dev eth0
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP mode DEFAULT group default qlen 1000
    link/ether 06:52:97:a0:56:be brd ff:ff:ff:ff:ff:ff

[niusmallnan]

When syslog info is used in the hook, it will call the function in /lib/dhcpcd/dhcpcd-run-hooks.

# Write a syslog entry
syslog()
{
	local lvl="$1"

	if [ "$lvl" = debug ]; then
		${syslog_debug} || return 0
	fi
	[ -n "$lvl" ] && shift
	[ -n "$*" ] || return 0
	case "$lvl" in
	err|error)	echo "$interface: $*" >&2;;
	*)		echo "$interface: $*";;
	esac
	if type logger >/dev/null 2>&1; then
		logger -i -p daemon."$lvl" -t dhcpcd-run-hooks "$interface: $*"
	fi
}

but the busybox logger doesn't support these params:

[root@ip-172-31-14-228 rancher]# logger -i
logger: invalid option -- 'i'
BusyBox v1.27.2 () multi-call binary.

Usage: logger [OPTIONS] [MESSAGE]

Write MESSAGE (or stdin) to syslog

	-s	Log to stderr as well as the system log
	-t TAG	Log using the specified tag (defaults to user name)
	-p PRIO	Priority (numeric or facility.level pair)

[rootwuj]

Tested with rancher/os:v1.5.2-rc1 with AWS from May 17
This version of the network service has no error log.
Verified fixed.