ntpviz - Don't plot a line during data absence

Gary E. Miller gem at rellim.com
Mon Oct 24 18:08:58 UTC 2016


Yo Hal!

On Mon, 24 Oct 2016 00:35:46 -0700
Hal Murray <hmurray at megapathdsl.net> wrote:

> >> It will also get occasional bogus breaks with the default maxpoll
> >> of 1024 when the peer logic filters out several samples.  
> 
> > Not the current logic.  As long as the peerstats stride is 1024 or
> > less the data is plotted as 'with lines'.  
> 
> You are not paying attention.  (or I'm doing a really crappy job of 
> explaining)

Or you misunderstand my objections.
 
> rawstats has a line for each response packet.

ntpviz does not use rawstats.  My guess is it never will.

> There is a filter between rawstats and peerstats.  Except for
> refclocks, most of the data gets filtered out.  So if the polling
> interval ramps up to 1024 seconds, there will be many places in
> peerstats where the spacing is over 1024.

Yeah, you keep saying that, but I keep not believing you.  I have 
never seen that behavior.  Got some logs that show that.

> Broken record time again.  If you set up a system to use the pool
> (and use the default maxpoll), you will have a test case to
> demonstrate that to you.

Broken record time, I don't care about the pool.

> Or you can use the data from one of your systems and split out the
> lines for a specific IP Address from peerstats and rawstats and run
> that into wc. $ wc -l glypnod-*.192.168.1.100
>    241 glypnod-peer.192.168.1.100
>    973 glypnod-raw.192.168.1.100
>   1214 total

Yeah, as expected.  I don't see how that gets to a poll over 1024
when maxpoll is 1024.

RGDS
GARY
---------------------------------------------------------------------------
Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
	gem at rellim.com  Tel:+1 541 382 8588
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 455 bytes
Desc: OpenPGP digital signature
URL: <http://lists.ntpsec.org/pipermail/devel/attachments/20161024/50a37432/attachment.bin>


More information about the devel mailing list