[Git][NTPsec/ntpsec][master] 4 commits: Tweak drift description.

[Gary E. Miller]

Gary E. Miller pushed to branch master at NTPsec / ntpsec


Commits:
4fc51b03 by Gary E. Miller at 2016-09-08T11:41:20-07:00
Tweak drift description.

- - - - -
09f59ff3 by Gary E. Miller at 2016-09-08T12:15:53-07:00
Add explanation for 'peer jitter'.  Tweak 'peer jitters' explanation.

- - - - -
6aa0c092 by Gary E. Miller at 2016-09-08T12:28:33-07:00
Add xplanation for peer offset

- - - - -
1c85a087 by Gary E. Miller at 2016-09-08T12:37:32-07:00
Tweak peer-offset, perr-offsets explannations.

- - - - -


1 changed file:

- ntpstats/ntpviz


Changes:

=====================================
ntpstats/ntpviz
=====================================
--- a/ntpstats/ntpviz
+++ b/ntpstats/ntpviz
@@ -746,13 +746,16 @@ larger frequency offsets.</p>
 """,
 
         "local-error": """\
-<p>This shows the frequency offset of the local clock.  It comes from
-field 3 of the loopstats log file.  The graph includes percentile data to
-show how much the frequency changes over a longer period of time.  The
-majority of this change should come from temperature changes (ex: HVAC,
-the weather, CPU usage causing local heating).</p>
+<p>This shows the frequency offset of the local clock (aka drift).
+It comes from field 3 of the loopstats log file.  The graph includes
+percentile data to show how much the frequency changes over a longer
+period of time.  The majority of this change should come from
+temperature changes (ex: HVAC, the weather, CPU usage causing local
+heating).</p>
+
+<p>Smaller changes are better.  An ideal result would be a horizontal
+line at 0ppm.</p>
 
-<p>Smaller changes are better.  An ideal result would be a horizontal line at 0ppm.</p>
 <p>Expected values of 99%-1% percentiles: 0.4ppm</p>
 """,
         "local-offset-histogram": """\
@@ -779,18 +782,39 @@ changes change the local clock crystal frequency and stability.  Crystals
 oscillate faster when warmer. This the single most important
 component of frequency drift.</p>
 """,
-        "peer-offsets": """\
-<p>This shows the offset of all remote clocks and refclocks.  This can
-be useful to see if offset changes are happening in a single clock or
-all clocks together.</p>
-<p>Closer to 0μs is better.  An ideal system would be a horizontal line
-at 0μs.</p>
+        "peer-jitter": """\
+<p>This shows the jitter of a remote peer or server.  Jitter is the
+current estimated dispersion; the variation in offset between samples.
+This is field 8 in the loopstats log file.</p>
+<p>Closer to 0s is better.  An ideal system would be a horizontal
+line at 0s.</p>
 """,
         "peer-jitters": """\
-<p>This shows the jitter of the remote clock offset.  In other words,
-how quickly the remote clock offset is changing.</p>
-<p>Closer to 0μs/s is better.  An ideal system would be a horizontal
-line at 0μs/s.</p>
+<p>This shows the jitter of all remote peers and servers.  Jitter is the
+current estimated dispersion; the variation in offset between samples.
+This is field 8 in the loopstats log file.</p>
+<p>Closer to 0s is better.  An ideal system would be a horizontal
+line at 0s.</p>
+""",
+        "peer-offset": """\
+<p>This shows the offset of a peer or server in seconds.
+This is field 5 from the loopstats log file.  This can be useful to
+see if the measured  offset is changing.</p>
+<p>The chart also plots offset±rtt.  Where rtt is the round trip time
+to the remote.  NTP can not really know the offset of a remote chimer,
+NTP computes it by subtracting rtt/2 from the offset.  Plotting the
+offset±rtt reverses this calculation to easier see the effects of
+rtt changes.</p>
+""",
+        "peer-offsets": """\
+<p>This shows the offset of all peers and servers in seconds.
+This is field 5 from the loopstats log file.  This can be useful to
+see if offset changes are happening in a single clock or all clocks
+together.</p>
+<p>Closer to 0s is better.  An ideal system would be a horizontal line
+at 0s. Typical 90% ranges may be: local serial GPS 200 ms; local LAN peer
+80µs; local PPS 20µs.  90% ranges for WAN servers may be 4ms and much larger.
+</p>
 """,
         }
 
@@ -951,8 +975,14 @@ ntpviz</a>, part of the <a href="https://www.ntpsec.org/">NTPsec project</a>
                     continue;
                 div_name = imagename.replace('-', ' ')
                 ifile.write("<div>\n<h2>%s:</h2>\n" % div_name)
-                if imagename in explanations:
-                    ifile.write("<div>\n%s</div>\n" % explanations[imagename])
+                if 'peer-jitter-' == imagename[:12]:
+                    exp = 'peer-jitter'
+                elif 'peer-offset-' == imagename[:12]:
+                    exp = 'peer-offset'
+                else:
+                    exp = imagename
+                if exp in explanations:
+                    ifile.write("<div>\n%s</div>\n" % explanations[exp])
                 gnuplot(image, os.path.join(outdir, imagename + ".png"))
                 div = imagewrapper % (imagename.replace(':', '%3A'), div_name)
                 ifile.write(div)



View it on GitLab: https://gitlab.com/NTPsec/ntpsec/compare/6a698051977f46ddcce7f7b641a773bac5d7e35a...1c85a087a6f92bbb7270b873ed32e1eec7fd4952
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ntpsec.org/pipermail/vc/attachments/20160908/26e83b91/attachment.html>