I have recently installed Slackware 13 (64bit) (Kernel 2.6.29.6) on an Acer Veriton M series with AMD Athlon cpu. I have started ntpd server and from it's log file it seems to be synchronising to each of the three stratum 2 servers specified in the ntp.conf file.

25 Mar 18:27:02 ntpd[4542]: time reset +2.163491 s
25 Mar 18:33:08 ntpd[4542]: synchronized to 130.88.200.4, stratum 2
25 Mar 18:37:25 ntpd[4542]: synchronized to 130.88.200.6, stratum 2
25 Mar 18:42:59 ntpd[4542]: time reset +2.282097 s
25 Mar 18:49:21 ntpd[4542]: synchronized to 130.88.200.6, stratum 2
25 Mar 18:50:17 ntpd[4542]: synchronized to 130.159.196.118, stratum 2
25 Mar 18:55:57 ntpd[4542]: synchronized to 130.88.200.6, stratum 2
25 Mar 18:58:57 ntpd[4542]: time reset +2.147508 s

If I invoke ntpdate from another linux machine, that ntpdate reports that the stratum of the ntpd server is too high and won't use it.

transmit(192.168.175.8)
receive(192.168.175.8)
transmit(192.168.175.8)
receive(192.168.175.8)
transmit(192.168.175.8)
192.168.175.8: Server dropped: strata too high
server 192.168.175.8, port 123
stratum 16, precision -20, leap 11, trust 000
refid [192.168.175.8], delay 0.02577, dispersion 0.00000
transmitted 4, in filter 4
reference time: cf562778.a310c705 Thu, Mar 25 2010 18:18:32.636
originate timestamp: cf562cdd.048cbee3 Thu, Mar 25 2010 18:41:33.017
transmit timestamp: cf562cd0.56618ce2 Thu, Mar 25 2010 18:41:20.337
filter delay: 0.02579 0.02577 0.02579 0.02579
0.00000 0.00000 0.00000 0.00000
filter offset: 12.68024 12.68024 12.68024 12.68024
0.000000 0.000000 0.000000 0.000000
delay 0.02577, dispersion 0.00000
offset 12.680248

25 Mar 18:41:20 ntpdate[9683]: no server suitable for synchronization found

I have run ntpd on another platform with Slackware 12 (32bit) and never had any problems.

well the fact that the strata is 16 just says that the server is not synchronized at all. run "ntpq -pn" to show the current status of the local service.

And to add to that, if you run an ntpdate (there's a new command, and ntpdate is depreciated now..can't remember off the top of my head..), and synchronize with a reference server, then start NTP, it'll come up. If the local clock is too far off, NTP won't start and/or it'll take a LONG time to step to the correct time. Forcing a sync, then starting the server, should make it come up.

Yes, agreed but the local clock was only a few seconds off when I started ntpd and is was started with the -g switch which should cope with a large initial offset. However I did try exactly what you suggest but it made no difference. But there are some more interesting results - see below.

This is the result of "ntpq -pn"

root@lunnahouse:~# ntpq -pn
remote refid st t when poll reach delay offset jitter
==============================================================================
130.88.200.4 193.62.22.66 2 u 34 64 17 54.605 575.592 323.068
130.159.196.118 192.93.2.20 2 u 48 64 17 58.951 543.223 324.398
130.88.200.6 193.62.22.98 2 u 20 64 17 55.302 163.562 321.107

These offset and jitter values look quite large and presumably ntpd considers them too large to offer itself as a credible server. The offset values can be as large as about 2000 (2 secs). It doesn't matter how long ntpd is left running, they never get small and stay small.

I have another identical machine with the same installation and the results from that are the same. There has been another development see below.

It is easier to read tabulated output in code tags (that's a link to instructions or you may prefer to use "Advanced Edit" mode which has a # button for code tags) like this

I decided to try ntpd on another platform - Xeon dual core with Slackware 12 (32bit kernel 2.6.21.5) and using the same ntp.conf. running ntpq -pn gives:

remote refid st t when poll reach delay offset jitter
==============================================================================
+130.88.200.4 193.62.22.66 2 u 57 64 377 54.244 5.643 0.615
+130.159.196.118 192.93.2.20 2 u 50 64 377 58.632 5.930 0.494
*130.88.200.6 193.62.22.98 2 u 20 64 377 54.250 7.204 0.405

These offset and jitter values don't change much anytime after about 15 mins from starting ntpd. This looks more normal. Running ntpdate (I will investigate what has replace ntpdate) from a another machine produces:

root@Marconi:~# ntpdate -d 192.168.175.12
27 Mar 07:45:54 ntpdate[22198]: ntpdate 4.2.4p0@1.1472-o Mon May 7 05:15:03 UTC 2007 (1)
Looking for host 192.168.175.12 and service ntp
host found : wallace.EsplLocal
transmit(192.168.175.12)
receive(192.168.175.12)
transmit(192.168.175.12)
receive(192.168.175.12)
transmit(192.168.175.12)
receive(192.168.175.12)
transmit(192.168.175.12)
receive(192.168.175.12)
transmit(192.168.175.12)
server 192.168.175.12, port 123
stratum 3, precision -20, leap 00, trust 000
refid [192.168.175.12], delay 0.02573, dispersion 0.00000
transmitted 4, in filter 4
reference time: cf58357b.f2ae33fe Sat, Mar 27 2010 7:42:51.947
originate timestamp: cf58362a.8c256940 Sat, Mar 27 2010 7:45:46.547
transmit timestamp: cf583632.cf8055fb Sat, Mar 27 2010 7:45:54.810
filter delay: 0.02574 0.02573 0.02573 0.02573
0.00000 0.00000 0.00000 0.00000
filter offset: -8.26316 -8.26317 -8.26317 -8.26317
0.000000 0.000000 0.000000 0.000000
delay 0.02573, dispersion 0.00000
offset -8.263170

27 Mar 07:45:54 ntpdate[22198]: step time server 192.168.175.12 offset -8.263170 sec

So this server is working. It anounces itself as a stratum 3 server which makes sense since it is feeding from stratum 2 servers.

So, where is the problem with the 64bit Slackware 13 installation? Is it a generic problem with the Acer motherboard or with the build of the kernel and/or ntpd? Don't know.

Many thanks to those who responded. If I make further discoveries, I will post them here.