Local ntp-server's strata too high (does not synch?)

[Thiemo]

Hi,

I am trying to set up a ntpd server for my lan. However, I cannot make the clients use it. I found that the strata of my ntp server is too high.

Code: Select all

ntpdate -vd ntp.thiam.ch
14 Aug 23:10:45 ntpdate[17881]: ntpdate 4.1.2@1.892 Tue Aug 12 00:54:43 CEST 2003 (1)
transmit(192.168.0.253)
receive(192.168.0.253)
transmit(192.168.0.253)
receive(192.168.0.253)
transmit(192.168.0.253)
receive(192.168.0.253)
transmit(192.168.0.253)
receive(192.168.0.253)
transmit(192.168.0.253)
192.168.0.253: Server dropped: strata too high
server 192.168.0.253, port 123
stratum 16, precision -16, leap 11, trust 000
refid [0.0.0.0], delay 0.02573, dispersion 0.00000
transmitted 4, in filter 4
reference time:    00000000.00000000  Thu, Feb  7 2036  7:28:16.000
originate timestamp: c2e67955.75622813  Thu, Aug 14 2003 23:10:45.458
transmit timestamp:  c2e67955.b55f56a7  Thu, Aug 14 2003 23:10:45.708
filter delay:  0.02599  0.02574  0.02573  0.02573
         0.00000  0.00000  0.00000  0.00000
filter offset: -0.24993 -0.25002 -0.25002 -0.25002
         0.000000 0.000000 0.000000 0.000000
delay 0.02573, dispersion 0.00000
offset -0.250023

14 Aug 23:10:45 ntpdate[17881]: no server suitable for synchronization found

I read at http://www.redhat.com/archives/rhl-beta ... 00027.html
that "starta too high" is because my server does not get synched. So I checked the log file after a restart and apparingly it does not synch.

Code: Select all

nyffeltrach root # /etc/init.d/ntpd restart && date && sleep 5 && tail -n 30 /var/log/ntpd.log
 * Stopping ntpd...                                                                                            [ ok ]
 * Initializing clock via ntpdate...                                                                           [ ok ]
 * Starting ntpd...                                                                                            [ ok ]
Don Aug 14 23:21:33 CEST 2003
14 Aug 23:20:11 ntpd[12293]: server returns a permission denied error
14 Aug 23:20:11 ntpd[12293]: server returns a permission denied error
14 Aug 23:20:11 ntpd[12293]: server returns a permission denied error
14 Aug 23:20:11 ntpd[12293]: server returns a permission denied error
14 Aug 23:20:11 ntpd[12293]: server returns a permission denied error
14 Aug 23:20:11 ntpd[12293]: server returns a permission denied error
14 Aug 23:20:11 ntpd[12293]: server returns a permission denied error
14 Aug 23:20:36 ntpd[12288]: ntpd exiting on signal 15
14 Aug 23:20:38 ntpd[12343]: frequency initialized 0.000 from /var/lib/misc/ntp.drift
14 Aug 23:20:38 ntpd[12348]: signal_no_reset: signal 17 had flags 4000000
14 Aug 23:20:40 ntpd[12348]: server returns a permission denied error
14 Aug 23:20:40 ntpd[12348]: server returns a permission denied error
14 Aug 23:20:40 ntpd[12348]: server returns a permission denied error
14 Aug 23:20:40 ntpd[12348]: server returns a permission denied error
14 Aug 23:20:40 ntpd[12348]: server returns a permission denied error
14 Aug 23:20:40 ntpd[12348]: server returns a permission denied error
14 Aug 23:20:40 ntpd[12348]: server returns a permission denied error
14 Aug 23:20:40 ntpd[12348]: signal_no_reset: signal 14 had flags 4000000
14 Aug 23:20:41 ntpd[12293]: parent died before we finished, exiting
14 Aug 23:21:32 ntpd[12343]: ntpd exiting on signal 15
14 Aug 23:21:33 ntpd[12397]: frequency initialized 0.000 from /var/lib/misc/ntp.drift
14 Aug 23:21:33 ntpd[12402]: signal_no_reset: signal 17 had flags 4000000
14 Aug 23:21:36 ntpd[12402]: server returns a permission denied error
14 Aug 23:21:36 ntpd[12402]: server returns a permission denied error
14 Aug 23:21:36 ntpd[12402]: server returns a permission denied error
14 Aug 23:21:36 ntpd[12402]: server returns a permission denied error
14 Aug 23:21:36 ntpd[12402]: server returns a permission denied error
14 Aug 23:21:36 ntpd[12402]: server returns a permission denied error
14 Aug 23:21:36 ntpd[12402]: server returns a permission denied error
14 Aug 23:21:36 ntpd[12402]: signal_no_reset: signal 14 had flags 4000000
nyffeltrach root #

I put all my servers defined in /etc/ntp.conf

Code: Select all

erver ntp1.tuxfamily.net               # 80.67.177.2
server ntp.univ-lyon1.fr                # 134.214.100.6
server ntp2.belbone.be                  # 195.13.1.153
server ntp.karpo.cz                     # 217.11.227.68

# Unknown type
server time.euro.apple.com              # 194.151.19.93
server ntp.unizar.es                    # 155.210.1.2
server hora.usc.es                      # 193.144.75.20

restrict 192.168.0.0 mask 255.255.255.0 notrust nomodify notrap
restrict default notrust nomodify

into /etc/conf.d/ntp NTPDATE_OPTS

Code: Select all

NTPDATE_WARN="y"
NTPDATE_CMD="ntpdate"
NTPDATE_OPTS="-b time.euro.apple.com"

but it was fruitless. That ntp does not work with these servers is very strange, for me, as

Code: Select all

nyffeltrach misc # ntpdate -vd time.euro.apple.com
14 Aug 23:42:09 ntpdate[12449]: ntpdate 4.1.1b@1.829 Sat Apr  5 23:33:31 CEST 2003 (1)
transmit(17.72.133.42)
receive(17.72.133.42)
transmit(17.72.133.42)
receive(17.72.133.42)
transmit(17.72.133.42)
receive(17.72.133.42)
transmit(17.72.133.42)
receive(17.72.133.42)
transmit(17.72.133.42)
server 17.72.133.42, port 123
stratum 2, precision -17, leap 00, trust 000
refid [17.72.133.55], delay 0.17586, dispersion 0.00084
transmitted 4, in filter 4
reference time:    c2e68074.635a426b  Thu, Aug 14 2003 23:41:08.388
originate timestamp: c2e680b1.dc1ffb48  Thu, Aug 14 2003 23:42:09.859
transmit timestamp:  c2e680b1.a8ae1049  Thu, Aug 14 2003 23:42:09.658
filter delay:  0.17586  0.18105  0.17621  0.17940
         0.00000  0.00000  0.00000  0.00000
filter offset: 0.124330 0.123505 0.122997 0.124026
         0.000000 0.000000 0.000000 0.000000
delay 0.17586, dispersion 0.00084
offset 0.124330

14 Aug 23:42:09 ntpdate[12449]: adjust time server 17.72.133.42 offset 0.124330 sec

seems to work perfectly well.

Does somebody know how to lower my strata, make my ntp server synch?

Cheers,

Thiemo

[lami]

After some searching I discovered this:
Your local ntp server needs to be synchronized with some other to lower it's stratum. So make sure it can connect to parent NTP server. You can use ntpq command to connect to local ntp server. Then use pe command to view status:

Code: Select all

diana root # ntpq
ntpq> pe
     remote           refid      st t when poll reach   delay   offset  jitter
==============================================================================
 adm.globe.cz    stratum.eunet.c  2 u   41   64    7    1.263   15.859  23.787
 mail.contactel. ntp0-rz.rrze.un  2 u   45   64    7    3.961   16.699  24.648

There should be something other than zero in reach column and probably some non-zero values in delay, offset and jitter (like in code above).

When you can connect, you have to wait a while, until ntpd calibrates your clock - it needs to know difference over time, so it can adjust it continuously. It takes about 15 minutes to sync. Your stratum value will be lower after some time. You can check your current stratum value using rv command inside ntpq:

Code: Select all

ntpq> rv
status=c011 sync_alarm, sync_unspec, 1 event, event_restart,
version="ntpd 4.1.1b@1.829 Thu Jul 31 14:38:47 CEST 2003 (1)",
processor="i686", system="Linux2.4.20-gentoo-r5", leap=11, stratum=16,
precision=-18, rootdelay=0.000, rootdispersion=2.925, peer=0,
refid=0.0.0.0, reftime=00000000.00000000  Thu, Feb  7 2036  7:28:16.000,
poll=4, clock=c3105c69.8d1f2dc2  Mon, Sep 15 2003 17:42:33.551, state=1,
offset=0.000, frequency=0.000, jitter=0.004, stability=0.000

ntpq> rv
status=0644 leap_none, sync_ntp, 4 events, event_peer/strat_chg,
version="ntpd 4.1.1b@1.829 Thu Jul 31 14:38:47 CEST 2003 (1)",
processor="i686", system="Linux2.4.20-gentoo-r5", leap=00, stratum=3,
precision=-18, rootdelay=15.984, rootdispersion=30.309, peer=41332,
refid=adm.globe.cz,
reftime=c3105e36.9a7e5215  Mon, Sep 15 2003 17:50:14.603, poll=7,
clock=c3105ea8.266bb55a  Mon, Sep 15 2003 17:52:08.150, state=4,
offset=-5.197, frequency=-35.498, jitter=9.186, stability=5.073

First report was generated shortly after restarting ntpd, you can see stratum=16 (look at the end of 3rd line). Second one is after 10 minutes. Stratum is 3 now. Now your clients can sync it's clock using your server:

Code: Select all

root@asuka:~# ntpdate -vd diana
15 Sep 17:52:43 ntpdate[28987]: ntpdate 4.1.1a@1.791 Wed May 29 19:52:32 PDT 2002 (1)
transmit(10.0.0.2)
receive(10.0.0.2)
transmit(10.0.0.2)
receive(10.0.0.2)
transmit(10.0.0.2)
receive(10.0.0.2)
transmit(10.0.0.2)
receive(10.0.0.2)
transmit(10.0.0.2)
server 10.0.0.2, port 123
stratum 3, precision -18, leap 00, trust 000
refid [81.95.96.3], delay 0.02580, dispersion 0.00000
transmitted 4, in filter 4
reference time:    c3105eb7.9954da4c  Mon, Sep 15 2003 17:52:23.598
originate timestamp: c3105ecb.a6ef1bac  Mon, Sep 15 2003 17:52:43.652
transmit timestamp:  c3105ecb.8a4b33da  Mon, Sep 15 2003 17:52:43.540
filter delay:  0.02602  0.02586  0.02582  0.02580 
         0.00000  0.00000  0.00000  0.00000 
filter offset: 0.111748 0.111769 0.111761 0.111755
         0.000000 0.000000 0.000000 0.000000
delay 0.02580, dispersion 0.00000
offset 0.111755

15 Sep 17:52:43 ntpdate[28987]: adjust time server 10.0.0.2 offset 0.111755 sec

I guess your stratum will be one less than stratum of your "parent" ntp server, so it will be 3 in most cases.

[lami]

I missed part of your post, so most of my post was useless. You have to allow access from other servers. I am not sure how exactly say: "these server are my parent, these are client and rest go away", but this is my working configuration:

Code: Select all

server ntp.globe.cz
server ntp1.contactel.cz

logfile         /var/log/ntpd.log
driftfile       /var/lib/misc/ntp.drift

restrict default nomodify
restrict ntp.globe.cz
restrict ntp1.contactel.cz
restrict 10.0.0.0 mask 255.255.255.0
restrict 127.0.0.1

[Thiemo]

Briallant!

Thanks lami. You helped me a lot. Both of your postings were necessairy.

I actually almost had given up.

Cheers

Thiemo