Joel, Jeff, Juergen - As stated, the intent of this draft is to specify in a logging-agnostic manner the minimum mandatory elements for I2RS traceability. I think we all agree this has real value and I don't view it as an obstacle in getting this work done (IMO in I2RS WG).
It seems the sticking point seems to be around the need for data modeling. To be clear, there was no intent to constrain or specify a preferred logging format or protocol. In fact, this was intended to be similar to the model used for SIP CLF where a format-agnostic data model (RFC6872) can be accompanied by one or many format-specific specifications (RFC6873 as an example), as deemed necessary by the community. The original ABNF and subsequent YANG was simply to provide a generic modeling of the mandatory logged elements that were described in the text with only the most minimal constraints possible imposed on character sets, etc based on the I2RS architecture. This was aimed at providing operation simplicity for implementors. If this goal is not met or if it is mis-interpreted as anything other than a logging format agnostic model, then it violates our prime directive and we can do one of three things: 1. Add supplementary clarifying text to the data modeling section explaining the intent described above. 2. If YANG is perceived as having more of a NETCONF bias we can go back to ABNF that is perhaps more universally viewed as protocol-agnostic. 3. Remove any data modeling elements from the draft entirely (ABNF or YANG or otherwise). We're as interested as anyone in making sure this work remains logging format and protocol agnostic, so we will happily comply with the preferred WG direction. To this point, could we perhaps adopt this work in the I2RS WG (presumably with all modeling removed) as a basis for defining the minimum mandatory elements needed for a traceability model for I2RS? Thanks, Gonzalo On Jun 10, 2014, at 3:58 PM, Joel M. Halpern <[email protected]> wrote: > Yes, I would agree that if the YANG model is dropped the document is both > useful and sufficiently clear. > > Yours, > Joel > > On 6/10/14, 3:34 PM, Jeffrey Haas wrote: >> On Tue, Jun 10, 2014 at 03:02:26PM -0400, Joel M. Halpern wrote: >>> In my personal view, it is clearly outside of our remit to reinvent >>> logging, no matter whether we like it or not. >> >> Agreed. Not in-scope for I2RS currently. >> >> Where would you suggest such work be done? >> >>> As such, the structural specificity of a YANG model (or ABNF) simply >>> does not belong in this document. The clarity about what fields >>> will be logged is needed. >> >> Would you agree that if the yang module is dropped that the info model is >> sufficiently clear for I2RS purposes? >> >> -- Jeff >> _______________________________________________ i2rs mailing list [email protected] https://www.ietf.org/mailman/listinfo/i2rs
