Re: Docsis subscriber management mib -13, submit to IESG?
Harrie Hazewinkel <[email protected]>
| Newsgroups | gmane.ietf.ipcdn |
|---|---|
| Message-ID | <[email protected]> |
On Wednesday, January 21, 2004, at 12:39 AM, Wilson Sawyer wrote: > Harrie - thank you. > > w.r.t. your first and third points, I'll add to or revise the wording. > > Do you have suggested text for your second point? I had chosen to err > on the > side of reticence but am willing to expand on it if you have something > in > mind. I have no real text and considering your comment it maybe is better this way. I can immagine that in the future filtering and, for instance, metering will be done. So explicitly forbidden it would not be a good idea. > As to your point (4), the short answer is "nothing". The cmFilterTable > is > unaffected, as are address entries other than "learned". Is there text > which > led you to believe otherwise? If you're referring to > docsSubMgtCpeControlActive=false, then the answer is similarly > "nothing" - > the docsSubMgtCmFilterEntry augments docsIfCmtsCmStatusEntry, and thus > is > bound to its existence. The FilterGroupTable applies to groups of > modems, > and is thus unaffected by the status of any particular modem. Yes, you are right. The augment takes care of of the existance of an entry for a particular modem and the FilterGroupTable is not affected by that. Correct me if I am wrong, but the filter group could be seen as an available configuration. Whether it is used or not is not a problem. This is similar to the approach of the DIFFSERV-CONFIG-MIB, only here you point to a specific sort of datapath element, the classifier element. Due to this I maybe found a 'bug', not sure if this is really true. In your MIB you always refer to the first pass of an classifier element, a diffServClfrElementEntry. However, the DIFFSERV-MIB also has a diffServClfrTable. This table is the common index of the complete classifier, not jsut an element. I beleive the difServClfrTable would be the right one to point to. Please read the commented 'text' above the diffServClfrTable of the DIFFSERV-MIB. I beleive that reflects this thought. Harrie