Wrong server time
tssync is still installed but we use it only if necessary.
we have "lost" the .c file of this module because we had changed services hosting a lot of month ago. We cannot find it cause anope site is down and we have found it into an old home directory backup.
...
Btw, now the situation is normalized. We see all channells from all servers...
i dont understand why...
we have "lost" the .c file of this module because we had changed services hosting a lot of month ago. We cannot find it cause anope site is down and we have found it into an old home directory backup.
...
Btw, now the situation is normalized. We see all channells from all servers...
i dont understand why...
-
TigerKatziTatzi
- Posts: 36
- Joined: Fri Apr 08, 2005 12:10 pm
my expierences after starting with unreal recently for a network with 5k daily max. running anope services.
- 18 servers
- all servers running ntpd.
- easiest way for time update => crontab 1x a day root ntpdate -u timeserverURL
- more complicated way ntpd running with perl script for several timeservers as reference. in case one of them is down
- running every two hours /os tssync , otherwise ur gaining too high offsets
- 18 servers
- all servers running ntpd.
- easiest way for time update => crontab 1x a day root ntpdate -u timeserverURL
- more complicated way ntpd running with perl script for several timeservers as reference. in case one of them is down
- running every two hours /os tssync , otherwise ur gaining too high offsets
- check ur ulines, they need to be correct (case sensitive)- me, from my server, dont see #great_channel with 150 users on it
- me, from my server, see #little_channel without modes (but registrered) with 1 user.
- another server admin, from his server, dont see #great_channel but see #little_channel with 11 users.
-
TigerKatziTatzi
- Posts: 36
- Joined: Fri Apr 08, 2005 12:10 pm
here some code we are using on some boxes:
Code: Select all
# this config will only allow the synch of the clock
# but offer no time services otherwise.
# ntpd has to be run as root to synch the clock.
# check if it is correctly working with:
# ntpq -p
# deny-by-default policy
restrict default ignore
# syncronize with these time servers
restrict tick.usno.navy.mil mask 255.255.255.255 nomodify nopeer noquery notrap
restrict time.twc.weather.com mask 255.255.255.255 nomodify nopeer noquery notrap
restrict bonehed.lcs.mit.edu mask 255.255.255.255 nomodify nopeer noquery notrap
restrict time.keneli.org mask 255.255.255.255 nomodify nopeer noquery notrap
restrict ntp.ourconcord.net mask 255.255.255.255 nomodify nopeer noquery notrap
restrict time-ext.missouri.edu mask 255.255.255.255 nomodify nopeer noquery notrap
restrict 192.73.48.2 nomodify mask 255.255.255.255 nopeer noquery notrap
server tick.usno.navy.mil
server time.twc.weather.com
server bonehed.lcs.mit.edu
server time.keneli.org
server ntp.ourconcord.net
server time-ext.missouri.edu
server 192.73.48.2
# local fudge if network servers not available
# server 127.127.1.0 # local clock
# fudge 127.127.1.0 stratum 10
restrict 127.0.0.1 nomodify
# track wander
driftfile /etc/ntp/drift
# logfile
logfile /var/log/ntp.log
Running ntpd is the way to go indeed (which you already do). Then, don't use anything else (like a services module), because that can only cause trouble.. :)
As for channels mysteriously not showing up in /list and all kinds of other weird stuff, yes.. that can happen if the time has changed (a lot) or is incorrect. For example, the creation time of channels >1d in the future would make it not show up in /list.
As for channels mysteriously not showing up in /list and all kinds of other weird stuff, yes.. that can happen if the time has changed (a lot) or is incorrect. For example, the creation time of channels >1d in the future would make it not show up in /list.