Serge, Can we table this unil the next snapshot? It certainly works as advertised here.
Dave Serge Bets wrote: > Hello David, > > On Sunday, March 23, 2008 at 20:07:33 +0000, David L. Mills wrote: > > >>You might peek with the ntpdc sysstats command and vet the counters. >>Verify no error messages at startup; the configuration code might have >>balked when attempting to lower the discard average. Verify with ntpq >>that the poll interval in fact starts out at 3. If the client does >>indeed start out at 3 and the server rate-limits it, use the restrict >>default limited kod command to see if a KoD comes back. > > > There are no configuration parsing errors: Both "minpoll 3" and > "average 3" are accepted (while "minpoll 2" or "average nan" shout the > expected error). > > By default sysstats shows no rate exceeded, with or without "average 3". > Rate exceeded start to appear when adding restrict limited without > "average 3". > > With restrict kod on the server, after some time the client logs "no > servers reachable" and doesn't poll anymore (but the refid stays "DCF", > no "RATE" or whatever). No problem with "average 3". > > The poll interval indeed starts at 3, and I see a pair of packets in and > out on the wire every 8-10 seconds. > > Anyway the kernel clock was never disciplined, even with "average 3" and > no limited/kod restrictions. Thanks Dave, I'll wait for the next > snapshot and retry. > > > Serge. _______________________________________________ questions mailing list [email protected] https://lists.ntp.org/mailman/listinfo/questions
