I use the following /etc/conf.d/ntp-client settings: NTPCLIENT_CMD="ntpd" NTPCLIENT_OPTS="--panicgate --slew --quit" This way i can re-use the settings from "ntpd" in /etc/ntp.conf. It works perfectly, but when it cannot connect to the ntp servers it does not seem to timeout and therefor blocks the boot process. The version I was using before was ntp-4.2.4_p6, this version did timeout so I wonder what changed in the code. When stracing I see the following: --- SIGALRM (Alarm clock) @ 0 (0) --- sigreturn() = ? (mask now []) clock_gettime(CLOCK_REALTIME, {1306425171, 670617738}) = 0 clock_gettime(CLOCK_REALTIME, {1306425171, 671257618}) = 0 sendto(19, "\343\0\6\364\0\0\0\0\0\0\0EINIT\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 48, 0, {sa_family=AF_INET, sin_port=htons(123), sin_addr=inet_addr("94.75.205.140")}, 16) = 48 clock_gettime(CLOCK_REALTIME, {1306425171, 672426486}) = 0 select(20, [16 17 18 19], NULL, NULL, NULL) = ? ERESTARTNOHAND (To be restarted) --- SIGALRM (Alarm clock) @ 0 (0) --- sigreturn() = ? (mask now []) clock_gettime(CLOCK_REALTIME, {1306425172, 670561022}) = 0 select(20, [16 17 18 19], NULL, NULL, NULL) = ? ERESTARTNOHAND (To be restarted) --- SIGALRM (Alarm clock) @ 0 (0) --- sigreturn() = ? (mask now []) clock_gettime(CLOCK_REALTIME, {1306425173, 670684700}) = 0 select(20, [16 17 18 19], NULL, NULL, NULL) = ? ERESTARTNOHAND (To be restarted) --- SIGALRM (Alarm clock) @ 0 (0) ---
It seems you are making the boot process depend on a good network connection. This means it is highly likely that the boot process stalls when there is no good network connection. Looking at the ntp-client conf.d script, it seems you had been warned: # Command to run to set the clock initially # Most people should just leave this line alone ... # however, if you know what you're doing, and you # want to use ntpd to set the clock, change this to 'ntpd' NTPCLIENT_CMD="ntpdate" If you want to really get this fixed as a bug, then ask at <http://www.ntp.org/>.
