In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] ("Ehrentraut, Diana") writes:
>
> I atached the graph I get. As you can see, the 29th of october is marked
> twice. It was the day of the daylightsavingtime(?) change.
> The strange thing about this is, that until the 31st of october I didn't
> have any problem with it, only in november.
It is indeed the daylightsaving change. When drawing the x-grid rrdtool
steps through the time by increasing the tm_hour field in the tm
structure. The time is initialized to 00:00:00. Now when rrdtool adds
one day to Oct 29 00:00:00 mktime(3) changes the day and hour value in
the tm structure to be 29 resp. 23 (i.e. Oct 30 00:00:00 with daylightsaving
set is Oct 29 23:00:00 without daylightsaving). I added debug output to
rrdtool which gives
find_next_time -> 972684000 / Sat Oct 28 00:00:00 2000
find_next_time -> 972770400 / Sun Oct 29 00:00:00 2000
find_next_time -> 972856800 / Sun Oct 29 23:00:00 2000
find_next_time -> 972943200 / Mon Oct 30 23:00:00 2000
find_next_time -> 973029600 / Tue Oct 31 23:00:00 2000
You can see the jump between line 2 and 3.
The question is now what would be the correct behaviour. 24 hours after
Oct 29 00:00:00 is Oct 29 23:00:00 and the graph shows the value for
this time/date. Maybe it should be rounded up to Oct 30 00:00:00.
Now what's the Right Behaviour?
> <<...OLE_Obj...>>
Nice Try ...
Rainer
--
[EMAIL PROTECTED]
--
Unsubscribe mailto:[EMAIL PROTECTED]
Help mailto:[EMAIL PROTECTED]
Archive http://www.ee.ethz.ch/~slist/rrd-users
WebAdmin http://www.ee.ethz.ch/~slist/lsg2.cgi
