Browse Prior Art Database

Optimized Process for Policy Information Collection Disclosure Number: IPCOM000018613D
Original Publication Date: 2003-Jul-28
Included in the Prior Art Database: 2003-Jul-28
Document File: 2 page(s) / 6K

Publishing Venue



In a policy system supporting solicited decisions, the "decision point" (the entity requesting the decision) can pass information to the "decision maker" (the entity making the decision). Such information guides the decision maker's decision processing. However, the information that needs to be provided depends on the current policies. We disclose an optimal mechanism for determining which data should be transmitted from decision point to decision maker.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 66% of the total text.

Page 1 of 2

Optimized Process for Policy Information Collection

In a Policy Based System, policy evaluations can either be solicited or unsolicited.

Unsolicited policy decision are delivered to a "decision point" (DP) from a "decision maker" (DM) without a DP explicitly requesting delivery. A policy system loading a new routing table is an example. Such decisions are addressed by such standards at the IETF/DMTF's PCIM and WSLA.

A solicited decision is requested by a DP from a DM. Examples include "can I (a resource) shutdown now" or "what discount should I apply to this purchase"? The problem is that the DP doesn't know what policies exist at the DM, so it doesn't know what to send.

For efficiency, in a solicited decision, the DP can pass information (parameters) to the DM on the request. In the current art, DPs either pass:
1. Nothing: the DM is expected to call back to the DP for required information.
2. Everything: the DP sends everything the DM might possible need.
3. A static set: the DP is configured to send some set of information. If that static set changes, the DP needs to be recompiled, or the DM must make a subsequent query to the DP.

These three approaches have drawbacks. 1 and 2 are inefficient, and 3 is either inefficient or fragile.

We disclose a dynamic process where the DM dynamically generates the list of relevant variables, and the DP downloads the list and sends the appropriate data. The act of sending only the appropriate data removed the inefficiencies out...