[collectd] Hit counter in threshold

[Florian Forster]

Hi Peter,

On Sun, Jul 12, 2009 at 11:39:02AM +1000, Peter Bray wrote:
> Just a small comment, hysteresis as used in the satellite control
> systems I've worked with over the last eight years, ...

so, what we do here *is* rocket science? Woohoo! ;)

> ... means N samples above/below a threshold to trigger the condition.
> So for example: if RH (Red High) >= 10.0 with hysteresis of 3, then
> you require at least three samples at or above 10.0 to trigger a red
> high alarm.

Okay, this is of course a nice and presumably [0] useful method. We
should definitely implement that, too, especially since it's easy to do.

What I meant was more along those lines: Take a value that can range
from zero to 100. Lets say values below 50 are good, values above 50 are
bad. To prevent flapping when the value is close to 50, set the state
depending on the previous state:
  if (value > 55) state = alarm
  else if (value < 45) state = okay
  else /* do not change state */

> This also seems to match the definition in the MacOSX Desktop
> Dictionary "the phenomenon in which the value of a physical property
> lags behind changes in the effect causing it".

Yeah, the term originates from magnetism or electricity, where the
reversion of an effect may need more or less energy than the energy
needed to cause the effect [1]. In a more mathematical sense one would
say that the value of a function depends on previous arguments (its
history).

> Others may have a different understanding of hysteresis but I thought
> I mention this understanding as naming is so important to
> understanding.

Right, using the generic term is probably not a good idea. Maybe the
method you described could be called “set hysteresis” while my
description would be “two-point hysteresis” (from the German term
„Zweipunktregler”, english: bang–bang control). Or do you know any
better terms?

Regards,
-octo

[0] People who shoot stuff into orbit and beyond try hard not to make
    mistakes ;)
[1] <http://en.wikipedia.org/wiki/Hysteresis#Magnetic_hysteresis>
-- 
Florian octo Forster
Hacker in training
GnuPG: 0x91523C3D
http://verplant.org/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
Url : http://mailman.verplant.org/pipermail/collectd/attachments/20090712/ccbd63d2/attachment.pgp