# Drift file (optional)
driftfile /var/ntp/ntp.drift
filegen peerstats file peerstats type day enable
filegen loopstats file loopstats type day enable
filegen clockstats file clockstats type day enable
remote refid st t when poll reach delay offset disp
==============================================================================
LOCAL(0) LOCAL(0) 3 l 20 64 377 0.00 0.000 10.01
clusternode1-pr 0.0.0.0 16 - - 1024 0 0.00 0.000 16000.0
*clusternode2-pr LOCAL(0) 4 u 528 1024 377 0.66 0.029 0.60
The asterisk indicates the prefered update node (where first to try to get the time).
remote is the address of the time server, with LOCAL(0) indicating the local clock.
refid indicated the type of the time server. LOCAL or .LCL. means local, .DCFa. is a DCF77 receiver, .PPS. is a hardware device generating a pulse every second.
st is the Stratum, which indicates the accuracy to be expected. Stratum 0 are usually atomic clocks, Stratum 1 might be radio controlled clocks.
when is the time to the next update.
poll is the count that when has to reach before an update is attempted
reach is an octal number that is left-shifted on each update. On a successful update an 1 is shifted in, 0 otherwise.
delay is the RTT to the time server
offset is the difference between the remote and the local clock
The way it's setup currently, your two servers only synchronise themselves locally and not to each other (which wouldn't be that good of an idea anyways without an additional good external source). Go to pool.ntp.org, pick a pool geographically close to you, and add it to your ntp.conf.
for some reason i will not go to synchronize the time with external NTP server.
now need a advise. As you said- two servers only synchronise themselves locally and not to each other.
I have two server are cluster mode and Oracle RAC is running
To change the time and date can I change it one after another in node 1 and Node2 (with date command)?? I don't want to restart the cluster and Oracle RAC.
will it be effect anything in cluster environment.
remote refid st t when poll reach delay offset disp
==============================================================================
LOCAL(0) LOCAL(0) 3 l 20 64 377 0.00 0.000 10.01
clusternode1-pr 0.0.0.0 16 - - 1024 0 0.00 0.000 16000.0
*clusternode2-pr LOCAL(0) 4 u 528 1024 377 0.66 0.029 0.60
Your problem is that "clusternode2-priv" cannot synchronise with "clusternode1-priv", as indicated by a reach of 0 (means no successful update during the last 8 attempts, each attempt 1024 seconds apart = about 2 hours 15 minutes). This can have several causes
No NTP server on the other node
No route to the other host
Unable to resolve the hostname to an IP address
I'd suggest changing the ntp.conf on node 1 to be the primary time source (no sync to node 2), and on node 2 to sync from node 1, synching every minute. That way you'll only have to change the time on node 1, and node 2 will catch up a minute later.
# Drift file (optional)
driftfile /var/ntp/ntp.drift
filegen peerstats file peerstats type day enable
filegen loopstats file loopstats type day enable
filegen clockstats file clockstats type day enable