ntp hell: System clock runs twice as fast as it should do
I've got a problem with the software clock on my RH 9 box running kernel 2.4.21-lufs-030704. It all started when I noticed that it seemed to be running more than twice as fast as it should, so I adjusted the drift to -500 to slow it down. This seemed to help, but it's now just a bit closer to being twice as fast as it should be!
I know ntp is supposed to automatically compensate for big differences so I ran ntptime to find out what was going wrong and found the lines in bold a little worrying:
ntp_gettime() returns code 0 (OK)
time c3bb81be.eed6c000 Fri, Jan 23 2004 11:19:26.932, (.932964),
maximum error 335376 us, estimated error 16 us
ntp_adjtime() returns code 0 (OK)
modes 0x0 (),
offset 0.000 us, frequency -500.000 ppm, interval 4 s,
maximum error 335376 us, estimated error 16 us,
status 0x1 (PLL),
time constant 0, precision 1.000 us, tolerance 512 ppm,
pps frequency 0.000 ppm, stability 512.000 ppm, jitter 200.000 us,
intervals 0, jitter exceeded 0, stability exceeded 0, errors 0.
From the ntp how-to I know that I'm missing PPSSIGNAL from status, which means no PPS signal has been detected. My question is: What do I do about it??
Thanks in advance,
Last edited by jimieee; 01-23-2004 at 06:19 AM.