[ntp:questions] monitoring symmetric peers
kmag at islab.demokritos.gr
Wed Jul 14 07:49:12 UTC 2010
Dave Hart wrote:
> On Tue, Jul 13, 2010 at 09:22 UTC, Kostas Magkos
> <kmag at islab.demokritos.gr> wrote:
>> But what about symmetrical peering? I cannot verify this association.
>> For example, when I query the servers for their peers with ntpq (ntpq
>> -p) they give out a 'u' (in the 't' column) for each other. Shouldn't
>> there be a 's' code, as in symmetric peer?
> Please see http://bugs.ntp.org/1338. As noted there, the type column
> in the peers billboard (ntpq -p) was pretty screwed up prior to
> 4.2.7p14. So, yes, if you use a recent 4.2.7 (ntp-dev) tarball, there
> should be a "s" code. Your OS is using ntpd 4.2.2 circa June 2006, so
> in about four years, you should see your type column improve.
Thanks for the info. You've got me thinking, that maybe its time to give
up on the debian stock version...
What other options are there to monitor the state of symmetric peers?
How can I be sure the peering works?
Thanks in advance,
More information about the questions