OPS Area RMONMIB WG Meeting Minutes 45th IETF Oslo, Norway July 14, 1999 Minutes by Andy Bierman Review Material (A) draft-ietf-rmonmib-rmonprot-ref-01.txt (B) draft-ietf-rmonmib-rmonprot-mac-01.txt (C) draft-romascanu-rmonmib-iftopn-mib-00.txt (D) draft-bierman-dsmon-mib-01.txt (E) draft-ietf-rmonmib-hcrmon-06.txt (F) draft-ietf-rmonmib-rmon2hc-00.txt (G) draft-ietf-rmonmib-rmonfull-00.txt (H) draft-ietf-rmonmib-rmonhc-00.txt (I) draft-dietz-apmmon-mib-00.txt Agenda 1) WG Status 2) Completion of current WG charter final HC-RMON MIB (E) RMON-2 MIB with HC TopNRateBase objects (F) RMON MIB in SMIv2 for advancement (G) RMON MIB in SMIv2 with HC TopNRateBase object (H) 3) Presentation and discussion of new I-Ds Interface TopN MIB (C) DS-MON MIB (D) APMMON MIB (I) 4) Discussion of updated proposal for new WG charter Refine list of deliverables Present charter proposal to the Area Director for consideration Minutes 1) WG Status The WG Chair presented a status summary of the I-Ds before the WG. 1.1) SMON MIB The SMON MIB has been published as RFC 2613. Work on this document is completed. Some vendors are already implementing the SMON MIB, and it is possible an Interoperability Test for the SMON MIB will be held towards the end of this year (maybe in September or October?). This issue will be discussed further on the mailing list. 1.2) RMON Protocol Identifier documents The latest RMON Protocol Identifier documents (A) & (B) are based on review comments received from Bert Wijnen and Randy Bush. Some historic protocols and the macros with incomplete REFERENCE sections have been removed from (B). The documentation of syntax rules in (A) is now more precise. The 'rule' prohibiting more than one 'fragments' BIT to be set in the protocolDirParameters OCTET STRING has been removed from (A). The Protocol Macros I-D (B) will be updated to emphasize that the WG is moving the document to historical status, rather than attempting to keep the document current (as new protocols are published). The existing macros will remain an an implementation aid to developers. A WG Last Call for this new I-D, along with (A), will be issued ASAP in order to complete this work item. 2) Completion of the current WG Charter The WG Editor presented a summary of the 4 new I-Ds recently published, which address two important open work items: HC TopNRateBase issue RMON MIB SMIv2 conversion A WG Last Call will be issued for all 4 documents. If no strong objections are raised, the documents will be forwarded to the Area Director for publication consideration. The changes to the 4 new I-Ds were highlighted: 2.1) (E) draft-ietf-rmonmib-hcrmon-06.txt This is the updated HC-RMON MIB. The duplicate hostTopNControlRateBase object from the RMON MIB (RFC 1757) and the nlMatrixTopNControlRateBase and alMatrixTopNControlRateBase objects from the RMON-2 MIB (RFC 2021) have been removed. The WG discussed the absence of a 64-bit equivalent for the alarmValue object. The RMON alarmTable does not officially support 64-bit thresholding. It is possible to use 'delta' alarms and keep the polling interval short enough to keep the difference between the polled values less than 2 billion. The WG decided not to add 64-bit alarmTable support to the HC-MON MIB. 2.2) (F) draft-ietf-rmonmib-rmon2hc-00.txt This is the updated RMON-2 MIB. Enumerations for high capacity rate bases have been added to the nlMatrixTopNControlRateBase and alMatrixTopNControlRateBase objects. 2.3) (G) draft-ietf-rmonmib-rmonfull-00.txt This is the SMIv2 conversion of the RMON MIB. The entire document has been updated to conform to new MIB template, and the SMIv1 MIB structure has been converted to SMIv2 structure. A conformance section has been added which reflects the conformance intent of the original RMON MIB. The WG is strongly urged to read this document carefully, since many significant modifications and additions were made. 2.4) (H) draft-ietf-rmonmib-rmonhc-00.txt This is the updated RMON MIB. It is based on (G), but the hostControlTopNRateBase object contains additional enumerations for high capacity rate bases. The WG decided to remove the utilization calculations in this document, because they are specific to half-duplex 10 MBit Ethernet. Rather than provide formulas for every conceivable media type, this text will be removed. It has been a source of confusion for implementors. 2.5) RFC 1513 Advancement The WG Editor agreed to update the Token Ring RMON MIB (RFC 1513) in SMIv2 format. The updated document will be subject to a 2 week WG Last Call, and then forwarded to the Area Director for publication consideration. 3) Presentation and discussion of new I-Ds The WG has solicited proposals for work items for a new charter proposal. Some WG members have responded by writing I-Ds describing the problem and a proposing a solution. The WG heard brief presentations of by the authors of each I-D. 3.1) Interface TopN MIB (C) A summary of (C) was presented. There were some issues raised in the discussion: interfaceTopNObjectVariable: there are no rules or specifics for setting this object no way for an NMS to know what an agent will support no appropriate handling of sparse tables or access control no support for Counter64 based objects The WG agreed to discuss this draft further on the mailing list, and attempt to reach agreement on the proper scope of the problem and the proper solution approach. 3.2) DS-MON MIB (D) A summary of (D) was presented. There were some minor issues raised in the discussion: the 'ds' prefix will be changed to a 'dsmon' prefix the dsPdistControlCollectMode object does not seem to add much value for the 'netLayer' collect modes. The same statistics can be collected with the dsStatsTable. Since collection is only useful at the application layer, this object will be removed. description in timestamp objects needs to be clarified in the event or a system restart 3.3) APMMON MIB (I) A summary of (I) was presented. For a detailed summary of this MIB, refer to the APMMON MIB slides. There were some issues raised in the discussion: there was agreement that some sort of metrics directory is needed, like the RMON-2 protocol directory. The MIB objects needed for such a directory are still under debate. the statistical mechanisms presented in APMMON are designed to eliminate the need for preselected distribution bins. Not all WG members agreed that bin-style statistics are such a problem. The WG agreed to study the APMMON MIB proposal and discuss the issue in detail on the mailing list. 4) Discussion of updated proposal for new WG charter The WG discussed the new charter proposal, as it was posted to the WG mailing list prior to the meeting. The name of the first work item was changed from 'Protocol Performance Metrics' to 'Application Performance Measurement'. New charter text for this work item was proposed and accepted by the WG. An updated charter proposal will be posted to the WG mailing soon, for any final comments before it is forwarded to the Area Director.