[ntp:questions] NTP makes a time jump

unruh unruh at invalid.ca
Wed Jul 10 06:32:30 UTC 2013


On 2013-07-10, E-Mail Sent to this address will be added to the BlackLists <Null at BlackList.Anitech-Systems.invalid> wrote:
> unruh wrote:
>> That was the model he had.
>> A server which all of the clients were supposed to stay
>>  within 50ms of, no matter whether or not that server was
>>  connected to a time source or not, and no matter if,
>>  after a long disconnect, that server then reconnected to
>>  a source and found itself 3 min out.  It was for him of
>>  paramount importance that all clients and server be
>>  withing 50 ms of each other. Whether they showed true
>>  time was much less important.
>>
>> Of course if his interest had been in each keeping the
>>  best time, the rest of the clients be damned, then this
>>  is not a good procedure.
>>
>> ...
>>
>> Note that I would not advise him to put his network of
>>  servers and clients into the pool for others to use as
>>  a time source.
>
> I think if his server were to run off into the weeds,
>  the pool service would decrease his server's score
>  every poll, and eventually remove it from the pool,
>  when his servers score got below 10.

While the pool has some defenses, I still would not advise him to
contribute to the pool. The bahaviour is different from most ntpd
servers (ie it will slew even if minutes off "true time") What it does
when disconnected from the net is not too much of a worry since it
cannot contribute to the pool then anyway. One does not need the pool
populated by a bunch of garbage sources despite some of the defenses the
pool has against them.



>
>  In addition, clients using multiple servers would
>   eventually change his server to outlyer, then reject,
>   as the poll results showed a growing offset as his
>   server drifted away.
>
>  If he supplied multiple servers, from separate multitier
>   sources, that eventually all traced back to the single
>   server that was drifting into the weeds, it would be
>   more likely that someone could get e.g. four of his
>   servers with different RefID IPs, and not many other
>   servers (pool or otherwise), and then those unlucky
>   clients would drift away with his servers, until the
>   dispersion of his servers got too high, and the clients
>   rejected his servers and went for more servers from the pool.
>



More information about the questions mailing list