| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Comments:
<complianc> it is best to have a driftfile defined
<FKnight`> doesn't work
<complianc> and a server line
<Pfft> is this linux ?
<complianc> fknight
<skoal> I love comstrip.
<complianc> you are doing something very wrong
<complianc> ntpd works fine.
<skoal> I dont know why I didn't think of this sooner.
<Pfft> rpm -qf `rpm -qa|grep ntpd` | grep conf
<complianc> perhaps opening up udp 123 on your firewalls
<FKnight`> complianc: how exactly can I be doing something very wrong if all I did was add a server line and restart ntpd ?
<Pfft> oh not qf ql. . hehe who knows
<skoal> complianc: Can I just define an internal DC?
<complianc> if you aren't getting your time, you are using a wrong or non compliant server.
<FKnight`> you don't need to open up inbound ports to synchronize ntp. besides, ntpdate works, so the ports are fine
<skoal> I don't wanna **** with my stupid firewall.
<FKnight`> I'm using the the pool servers from ntp.org that work fine with ntpdate
<skoal> [root@aaa etc]# ntpdate
<skoal> 25 Mar 12:15:01 ntpdate[25933]: no servers can be used, exiting
<complianc> skoal, i don't know, where are your DCs getting their time from?
<skoal> DCs are going outside
<skoal> But I want this box to sync with the AD boxes.
<pale_blue> ntpdate servername
<complianc> do dc's run complant ntp software?
<Pfft> compliant enough
<skoal> yeah.
<peerce> microsoft uses a RPC thing for net time
<Pfft> ntp will talk with dc's just fine
<peerce> or at least, they used to.
<peerce> wasn't NTP
<Pfft> we do it all over the place
<complianc> then feel free to point at your dcs
<FKnight`> NET TIME != Windows Time Service
<peerce> they got a NTP client, but I don't think its a server
<skoal> ntp socket is in use!
<complianc> *tick.usno.navy. 209.51.191.20 1 u 275 1024 377 12.204 0.084 0.066
<complianc> -clock.fmt.he.ne 209.51.191.20 1 u 1016 1024 377 94.977 5.227 5.906
<complianc> +time.nist.gov 209.51.191.20 1 u 1006 1024 165 53.290 2.940 0.238
<complianc> +avi-lis.gw.ligh 209.51.191.20 1 u 952 1024 377 1.187 -0.801 0.300
<complianc> +bonehed.lcs.mit 209.51.191.20 1 u 7 1024 377 17.044 -0.565 0.010
<complianc> strat 1 baby.
<XTR> omg.
<skoal> DO i need to kill ntpd to use ntpdate?
<pale_blue> net time /setsntp:...
<Pfft> skoal - yes
<complianc> e-1203482108421482130480214802184 accuracy!
<complianc> of course, the oscillator in these **** pcs
<complianc> is inaccruate as hell
<skoal> 25 Mar 13:28:26 ntpdate[25949]: step time server 172.16.10.1 offset 4325.366770 sec
<skoal> HEE HEE
<complianc> ntpdate -g
<Aurelius> omg.
<complianc> to override such a big step
<complianc> i believe
<FKnight`> so what exactly am I doing wrong for ntpd not to sync? ntpdate <server works> so it's not the server or my fireall.
<FKnight`> *firewall
<complianc> anything > 1000ms won't do it
<peerce> at work, we have a GPS clock in one of our data centers on a Sparc/Solaris system with NTP, then all our Cisco routers run NTP synched off that, and everyone is supposed to use the nearest Cisco gateway as their NTP reference.
<complianc> unless you force it
<peerce> works quite well
<Aurelius> clock skew
<complianc> standard issue tiered ntp setup peerce
<XTR> Cock stew.
<skoal> -g?
<complianc> the key is the GPS clock
<skoal> that doesnt work
<Aurelius> server 192.43.244.18, stratum 1, offset -0.001153, delay 0.09285
<complianc> man ntpdate
<skoal> nope.
<skoal> no man page! :~((
<Aurelius> ****ing 1.2ms
<peerce> $ ntptrace
<peerce> localhost: stratum 3, offset 0.000010, synch distance 0.09323
<peerce> mewmie.mainecoon.com: stratum 2, offset -0.002382, synch distance 0.01445
<peerce> prowlingtabby.mainecoon.com: stratum 1, offset 0.002992, synch distance 0.00035, refid 'GPS'
<complianc> sorry, that's ntpd
<peerce> thats home.
<FKnight`> in order for me to sync my time, I have this in my cron: /usr/sbin/ntpdate north-america.pool.ntp.org . ntpd just refuses to work.
<Aurelius> lame *** ****
<complianc> ntpd is -g for > 1000ms of change
<skoal> I dont even have an ntpd man page.
<complianc> ntpdate should probably just do it
<skoal> OH wait.
<skoal> it worked.
<skoal> that offset took
<skoal> i didnt need to override ****
<Aurelius> ntpdate will set to whatever
<complianc> right, i was thiking ntpd
<complianc> not ntpdate
<skoal> you're a roxtar.
<complianc> ntpdate == for dummies
<complianc> ntpd == for real time.
<skoal> now to fix up ntp.conf
<Aurelius> complianc: i use ntpd
<skoal> comstrip here we come!
<peerce> well, I run ntpdate before I launch ntpd to preset the clock on a reboot.
<Aurelius> i use ntpdate to check it periodically :)
<FKnight`> ntpdate is for people who have copies of ntpd that are properly configured yet refuse to work.
<Aurelius> i only use ntpdate -q
<skoal> FKnight` AHHA
<complianc> must be the protocol fknight. clearing ntpd doesn't work.
<skoal> okay
<skoal> very good.
<skoal> ntp is fixed.
<skoal> thanks guys
<skoal> now i gotta watch this cron job.
<XTR> np dude.
<FKnight`> skoal what was wrong
<complianc> ntpq -c opeer | wc -l 36
<skoal> bull**** server entry
<Aurelius> he was a douchebag
<complianc> totally broke.
<peerce> yah, at boot, I run...
<FKnight`> what was the bull**** server entry ?
<skoal> ntpdate <internal server> fixed my huge offset
<skoal> and then I just added the internal server to the ntp.conf
<complianc> any invalid server will cause ntp headaches
<skoal> and started ntpd back up
<Aurelius> heh, i only use time.nist.gov
<Aurelius> :(
<peerce> ntpdate -b (pick-a-server); setclock; ntpd &
<skoal> [root@aaa etc]# /sbin/service ntpd start
<skoal> ntpd: Synchronizing with time server: [ OK ]
<skoal> Starting ntpd: [ OK ]
<Aurelius> i should add more
<complianc> if you are trying to set time with ntpd with an adjustment > 1000ms, ntpd will NOT do it.
<complianc> other than that, it's a reachabilty issue.
<complianc> -g Normally, ntpd exits if the offset exceeds the sanity limit,
<complianc> which is 1000 s by default.
<peerce> my server's cmos clock is ****ty, so I use that `ntpdate -b server; setclock; ntpd
<FKnight`> the only server entry I have that I think is bogus is "server 127.127.1.0" which the ntp.conf, which everyone tells me is self documenting, specifically states not to remove.
<complianc> hahaha
<complianc> can you ping 127.127.1.0?
<complianc> remove it. put in a real server.
<peerce> localhost is 127.0.0.0/8
<FKnight`> complianc: it says not to remove it.
<complianc> hahah
<pale_blue> wtf. what's the server line read EXACTLY right now? because when i start the service it doesn't say **** about syncrhonizing with server.
<complianc> THEN DON'T!
<complianc> hahahah
<peerce> i use...
<peerce> # some stratum 2 servers....
<peerce> server ntp.lmi.net
<peerce> server ntp1.sf-bay.org
<peerce> server ntp2.sf-bay.org
<peerce> server reloj.kjsl.com
<peerce> server ntp1.mainecoon.com
<Aurelius> peerce: but most stacks only listen for 127.0.0.1/32
<complianc> THEY TOLD ME NOT TO DO IT.
Return to
#windows
or
Go to some related
logs:
#heroin
#online
#beginner
#slackware
yum segmentation fault parsing package install arguments
#politics
#worldcup
ran_ver efnet
#politics
mpeg2 good bitrate
|
|
