Document: draft-ietf-tsvwg-tcp-mib-extension-12.txt Reviewer: Eric Gray Review Date: 11/22/2006 IETF LC End Date: 11/27/2006 Summary: This draft is not quite ready for publishing as a Proposed Standard. There are several relatively minor issues which should be addressed prior to publishing this draft. Comments: ++++++++ "this () implementation" ambiguity ================================== Starting with the second paragraph on page 11, and continuing through the remainder of section 3.2, you use the phrase "this implementation" several times in such a way that it is not clear either that you are talking about a specific TCP implementation, or which implementation that might be. For example, in the 2nd paragraph on page 11, "this specific implementation" seems meant to include "any specific TCP implementation" while, in the 3rd paragraph, "this TCP implementation" seems to refer to "a specific TCP implementation" examined by way of the MIB. The 4th paragraph opens by mentioning that the MIB might not be useful in comparing different implementations, and closes by saying an implementer may use the MIB and their own detailed knowledge of a specific TCP implementation (presumably one of the TCP implementations they've implemented) to debug/evaluate "this implementation" - seeming to refer to the same implementation of which they have detailed knowledge (as opposed to in comparison with other implementations). I suggest the following modifications: 2nd para - replace "this specific implementation" with "any specific TCP implementation" 3rd para - replace "this TCP implementation" with "TCP implementations" 4th para - replace "this implementation" with "their implementation" As a related question, is their a specific requirement stated in this specification that identifies exactly how a network management platform, network operator, or TCP implementer is expected to determine which TCP implementation the objects in the MIB are associated with - or is it merely assumed that they will be able to do this using either some other MIB objects, management information, or a priori knwoledge about the network and the specific device being queried? MIB Issue - flaws in descriptions ================================= DESCRIPTION at the top of page 16 (2nd para): should read "TCP Stack Statistics" as opposed to "TCP Path Statistics" (cut and paste error?) 2nd DESCRIPTION on page 16 (2nd para): should read "TCP Application Statistics" as opposed to "TCP Path Statistics" (cut and paste error?) 3rd DESCRIPTION on page 16 (2nd para): should read "TCP Tuning Statistics" as opposed to "TCP Path Statistics" (cut and paste error?) 1st paragraph on page 17: replace with appropriate text to reflect that this is an indicator of the active state of "notifications" for this MIB (as opposed to "TCP Path Statistics") NITs: ==== 4th sentence, 1st paragraph, section 3.2: "... that TCP has to solve ..." as opposed to "... that TCP has to solved ..." Last sentence, 1st paragraph, page 11: "... TCP implementations ..." as opposed to "... TCP implementation ..." 1st sentence, 2nd paragraph, page 11: "Since the underlying algorithms are not uniform, it is difficult to tightly specify a MIB." ("it makes it" is redundant, since the first "it" apparently refers to the fact that underlying algorithms are not uniform, and this redundancy makes the phrasing awkward) Alternatively: "That fact that the underlying algorithms are not uniform makes it difficult to tightly specify a MIB." 2nd sentence, 1st paragraph, section 3.3: "... can be used to defend ..." as opposed to "... can be to defend ..." Last paragraph on page 12 is awkward. I suggest rewording as follows: "Most diversity in SYN flood defense arises from variations in these algorithms to limit load, and therefore cannot conveniently be instrumented with a common standard MIB." 1st DESCRIPTION on page 27: "whatever time granularity the system supports" as opposed to "what ever time granularity is the system supports."