Wrong Date and time

[myo38]

Jocko,

Sorry, I typed -AS instead of -A5 !

Here is the good result :

Code: Select all

# grep '<ntp>' /etc/nas_conf_db.xml -A5
                <ntp>
                        <enable>true</enable>
                        <ip1>ntp.midway.ovh</ip1>
                        <ip2>ntp.unice.fr</ip2>
                        <server>true</server>
                </ntp>

Michel

[Jocko]

Hi

So all seems ok :scratch

Then we try now to start (restart manually the service) :

Code: Select all

ps axf|grep ntpd
killall ntpd
ps axf|grep ntpd
php -f /direct-usb/fvdw/cron-boot/restart-sync-ntp-daemon.php

Here the script will wait at least 90'' (if you did not change the sleep value). And when the script is completed, do again

Code: Select all

ps axf|grep ntpd

[myo38]

Hi Jocko,

I did what you said. Here is the script :

root@Serveur-Maison:/ # ps axf|grep ntpd
2125 pts/0 D+ 0:00 \_ grep ntpd
1456 ? Ss 0:00 /bin/ntpd -l
1622 ? Ss 0:00 /bin/ntpd -p ntp.midway.ovh -p ntp.ensma.fr
root@Serveur-Maison:/ # killall ntpd
root@Serveur-Maison:/ # ps axf|grep ntpd
2134 pts/0 R+ 0:00 \_ grep ntpd
root@Serveur-Maison:/ # php -f /direct-usb/fvdw/cron-boot/restart-sync-ntp-daemon.php
root@Serveur-Maison:/ # ps afx|grep ntpd
2173 pts/0 S+ 0:00 \_ grep ntpd
2147 ? Ss 0:00 /bin/ntpd -p ntp.midway.ovh -p ntp.ensma.fr
root@Serveur-Maison:/ #

I hope it helps.

I just have to add that once a while my server gets the good date and time. I don't know if it's due to the script (I've put a 300 sleep) or just that the restart of the fiber box is quick enough.

Michel.

[Jocko]

Hi

So no issue with thé script! But...

I just have to add that once a while my server gets the good date and time. I don't know if it's due to the script (I've put a 300 sleep) or just that the restart of the fiber box is quick enough.

I think there is an issue with the current max running Time value which aborts the script if you use this sleep value. I go to check what is thé current value.

[Jocko]

Hi

Nope it is not the issue...
The firmware uses 60" as max execution time

PhpMaxRunTime.jpg

So the script should abort every time... And indeed when running PHP from a command line (it is your case) the default setting is 0 (so disabled). Then no issue on this side.

However to confirm this point you may edit the script restart-sync-ntp-daemon.php :

Code: Select all

<?php
require_once("/etc/inc/global_defs.inc");
require_once("/etc/inc/globals.inc");
require_once("/etc/finc/logError.finc");
require_once("/etc/finc/configfunc_read.finc");
require_once("/etc/finc/ntpfunc_daemon_ntp_kill.finc");
require_once("/etc/finc/ntpfunc_daemon_start.finc");

#Reset running configuration
set_time_limit(0);

$config = "";
$volatileConfig = "";

if (($rc= configfunc_read($config, $volatileConfig)) != OK_E){
    logError(__FILE__,  __LINE__ , "nasapi: Error reading conf files. err num = {$rc}", LOG_WARNING);
    return $rc;
}

# kill all orphan daemon
ntpfunc_daemon_ntp_kill();

$config['system']['ntp']['enable'] = "true";

# Custom a delay in seconds before starting ntp daemon
sleep(300);

$rc =  ntpfunc_daemon_start();

if ($rc != OK_E){
    logError(__FILE__, __LINE__ , "ntpfunc_ntp_start, err= {$rc}", LOG_WARNING);
    return FAIL_E;
}

return OK_E;

?>

So I believe 300" (5min) may be not enough. Sometime a box may take a long time to be synchronized. then you can increase again the sleep value.

[myo38]

Jocko,

Thanks again for your help.

I did edit the script. And I have been happy : it worked the first day.
But then it seems to be working once every other day ! And now no more.

Probably that the fiber box started quickly enough to allow access to web time servers once a while.

I'm really disappointed !

What could be done ?

Michel.

[Jocko]

Hi

:thinking

Can you post a screen capture of "Boot jobs" menu and attach your current version of restart-sync-ntp-daemon.php

[myo38]

Jocko,

No problem, with pleasure !

Thanks again,
Michel

[Jocko]

Hi,

all still seems ok :scratch
Note: I would have prefer you post the file and not a capture when you edit it

Anyhow we go to make a manual test... for a first test set a sleep value 5" instead of 300
So do

Code: Select all

 ps axf |grep ntpd| grep -v 'grep'
date
killall ntpd
date -s 2019.01.01-12:00
date
debugon
php -f /direct-usb/fvdw/cron-boot/restart-sync-ntp-daemon.php
debugoff
cat /var/log/messages
ps axf |grep ntpd| grep -v 'grep'
(wait around 1min or repeat date command up to date is updated)
date


[myo38]

Hi,

Sorry, but neither a .php or a .txt or a capture file had been accepted. So I did how I could.

Here is the trace of the test you asked me :
# ps axf |grep ntpd| grep -v 'grep'
1456 ? Ss 0:00 /bin/ntpd -l
root@Serveur-Maison:/ # date
Mon Jan 1 21:22:52 CET 2018
root@Serveur-Maison:/ # killall ntpd
root@Serveur-Maison:/ # date -s 2019.01.014-12:00
Mon Jan 14 12:00:00 CET 2019
root@Serveur-Maison:/ # date
Mon Jan 14 12:00:05 CET 2019
root@Serveur-Maison:/ # debugon
root@Serveur-Maison:/ # php -f /direct-usb/fvdw/cron-boot/restart-sync-ntp-daemon.php
root@Serveur-Maison:/ # date
Wed Nov 16 15:26:50 CET 2022
root@Serveur-Maison:/ # debugoff
root@Serveur-Maison:/ # cat /var/log/messages
Jan 14 12:00:14 Serveur-Maison syslog.info syslogd started: BusyBox v1.21.0
Jan 14 12:01:34 Serveur-Maison authpriv.notice sudo: root : TTY=pts/0 ; PWD=/ ; USER=root ; COMMAND=/bin/ps -ef
Jan 14 12:01:39 Serveur-Maison authpriv.notice sudo: root : TTY=pts/0 ; PWD=/ ; USER=root ; COMMAND=/bin/ntpd -p ntp.midway.ovh -p ntp.ensma.fr
Nov 16 15:27:02 Serveur-Maison syslog.info syslogd exiting
root@Serveur-Maison:/ # ps axf |grep ntpd| grep -v 'grep'
2054 ? Ss 0:00 /bin/ntpd -p ntp.midway.ovh -p ntp.ensma.fr
root@Serveur-Maison:/ # date
Wed Nov 16 15:28:03 CET 2022

So it seems that the .php cron is working well ! For tomorrow I'll reduce the wait time to 59 sec, and I'll do it twice. We'll see.

Michel.