I've inherited a linux server that had NTP running happily as a client (syncs to some webserver), and is the machine that a windows domain looks to for syncing their clocks. Something seems to have gone awry in an upgrade and now windows machines get a mysterious error trying to sync to the machine:
"The time sample was rejected because: Packet contains unreasonable root delay or root dispersion values. This may be caused by poor network conditions."
So I go to NTP on the linux server, as a client it appears to be syncing correctly according to logs. However I go to /etc/init.d/ntp, and when I run status, says "unused", and running start gets me "no server configured".
I can't tell what machine or configuration is at fault because I thought ntp ran both as client and server at the same time. Ideas? (Will x-post to another subreddit after I clear away the linux side of the equation at fault)
Thanks a bunch. :) Edit: Formatting.
Edit: Thanks guys, I'll be back on this on monday, had to leave it until later.
[link] [9 comments]