From e4844b8a5fb0159281dbdbace333f69f6d3f1b4b Mon Sep 17 00:00:00 2001 From: Denys Vlasenko Date: Fri, 1 Jan 2010 21:59:49 +0100 Subject: [PATCH] ntpd: add a comment, no code changes Signed-off-by: Denys Vlasenko --- networking/ntpd.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/networking/ntpd.c b/networking/ntpd.c index d4754f8..02761b3 100644 --- a/networking/ntpd.c +++ b/networking/ntpd.c @@ -432,7 +432,7 @@ filter_datapoints(peer_t *p, double t) sum += dispersion(&p->filter_datapoint[idx], t) / (2 << i); if (minoff == p->filter_datapoint[idx].d_offset) { - minoff -= 1; + minoff -= 1; /* so that we don't match it ever again */ } else if (maxoff == p->filter_datapoint[idx].d_offset) { maxoff += 1; @@ -447,6 +447,13 @@ filter_datapoints(peer_t *p, double t) wavg += x; /* add another older6/64 to form older6/32 */ p->filter_offset = wavg; p->filter_dispersion = sum; +//TODO: fix systematic underestimation with large poll intervals. +// Imagine that we still have a bit of uncorrected drift, +// and poll interval is big. Offsets form a progression: +// 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7, 0.7 is most recent. +// The algorithm above drops 0.0 and 0.7 as outliers, +// and then we have this estimation, ~25% off from 0.7: +// 0.1/32 + 0.2/32 + 0.3/16 + 0.4/8 + 0.5/4 + 0.6/2 = 0.503125 // +----- -----+ ^ 1/2 // | n-1 | -- 2.7.4