[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

snmpconf Issue 2 - Split of filter and action into two objects


Some have argued that the split of filters and actions into
two different MIB objects is not correct.  This is related to
issue 1.

Again, http://www.escribe.com/computing/snmpconf/m275.html is a
useful thread to look at.

Juergen Schoenwaelder writes:  "this level of granularity
(two objects) will just turn out to be very impractical if
you express real-world policies."  Juergen Quittek argues
in a similar fashion that "[the separation] is impractical
because of the fine granularity. The problem is not only the
separation of filters and actions but also the implied separate
handling of each policy."  Instead, the argument is that the
filter/action pair should be exposed as a single MIB object
and that the Script MIB could then be used to transport them.

Others have argued that the split is very useful.  Steve
Waldbusser has given several reasons why he believes this is
true, including that policy filters and actions are conceptually
quite different and should be separate for modeling reasons.  In
adddition, he wrote:

- Because all filter information is separate, the
  manager can inspect the filter and understand whether
  to download the policy to an agent
- Because filter information is separate, the agent can schedule
  filter evaluation separately from action evaluation.
- The filter expression is known to be idempotent, information
  that both manager and agent can use for optimization.
- It is easier to re-use filter and action expressions when they
  are separate.