Fuel masters' ntp.conf contain settings for both Local Refclocks types
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Fuel for OpenStack |
Fix Committed
|
Medium
|
Stanislaw Bogatkin |
Bug Description
Existence of settings for both, Undisciplined Local Clock and Orphan Mode, can provide unstable ntp behavior.
For example unavailability of self-sync in case of isolated cluster:
~# ntpdate -d 192.168.1.5
19 Dec 18:32:09 ntpdate[8156]: ntpdate 4.2.6p3@1.2290-o Tue Jun 5 20:12:09 UTC 2012 (1)
Looking for host 192.168.1.5 and service ntp
host found : 192.168.1.5
transmit(
receive(
transmit(
receive(
transmit(
receive(
transmit(
receive(
transmit(
192.168.1.5: Server dropped: Server has gone too long without sync
server 192.168.1.5, port 123
stratum 7, precision -22, leap 00, trust 000
refid [192.168.1.5], delay 0.02605, dispersion 0.00005
transmitted 4, in filter 4
reference time: 00000000.00000000 Mon, Jan 1 1900 0:00:00.000
originate timestamp: d83eedb9.4138e3af Fri, Dec 19 2014 18:32:25.254
transmit timestamp: d83eedb9.40e33c69 Fri, Dec 19 2014 18:32:25.253
filter delay: 0.02614 0.02625 0.02605 0.02611
0.00000 0.00000 0.00000 0.00000
filter offset: 0.000778 0.000722 0.000841 0.000815
0.000000 0.000000 0.000000 0.000000
delay 0.02605, dispersion 0.00005
offset 0.000841
19 Dec 18:32:27 ntpdate[8156]: no server suitable for synchronization found
Changed in fuel: | |
status: | New → Confirmed |
importance: | Undecided → Medium |
assignee: | nobody → Stanislaw Bogatkin (sbogatkin) |
Changed in fuel: | |
status: | Confirmed → Triaged |
Changed in fuel: | |
status: | In Progress → Fix Committed |
As long as we don't use (and will not) master node in HA mode, I think that we should just to remove tos orphan string from ntp config.