RE: Last Call: 'Cable Device Management Information Base forData-Over-Cable Service Interface Specification CompliantCableModem s and Cable Modem Termination Systems' to ProposedStandard
"Jean-Francois Mule" <[email protected]> Wed, 21 Dec 2005 13:49:41 -0700
| Newsgroups | gmane.ietf.ipcdn |
|---|---|
| Message-ID | <[email protected]> |
Rich, Brian and all, Given where we are in the process and given the existing restrictions in RFC2669, my first reaction would be to do a real check with operators and implementers by asking: a) CM operators about their actual use of the syslog server DHCP option in field deployments (multi-value or single?). b) CM vendors about how their implementations deal with multi-values in the mentioned DHCP options. Given that syslog is UDP-based, what would it mean to have multiple server values? It can't reliably be used for fail-over scenarios so should mean broadcast syslog. I personally think that this would best be done by using a management station that could act like a trap exploder and broadcast or multicast the syslog msg to multiple servers. Comments and feedback from the list is appreciated. Jean-Francois > -----Original Message----- > From: Richard Woundy @ Comcast > Sent: Wednesday, December 21, 2005 1:01 PM > To: Brian Hedstrom; Wijnen, Bert (Bert) > Cc: Ipcdn (E-mail); Richard Woundy @ Comcast > Subject: RE: [ipcdn] Last Call: 'Cable Device Management Information > Base forData-Over-Cable Service Interface Specification > CompliantCableModem s and Cable Modem Termination Systems' to > ProposedStandard > > Brian, Bert, > > I need some help from the IPCDN working group to resolve this Last > Call > objection. > > The latest DOCSIS RFI specification, > <http://www.cablemodem.com/downloads/specs/CM-SP-RFI2.0-I10- > 051209.pdf>, > annex D.1.1, specifies the use of five standard DHCP options: > > * Option code 1 (Subnet Mask) > * Option code 2 (Time Offset) > * Option code 3 (Router Option) > * Option code 4 (Time Server Option) > * Option code 7 (Log Server Option) > > Looking at RFC 2132, <http://www.ietf.org/rfc/rfc2132.txt?number=2132>, > it becomes obvious that DHCP options 1 and 2 are single-valued options, > and DHCP options 3, 4 and 7 are potentially multi-valued options. > > So it would seem that if we are worried about multi-valued DHCP > options, > we should be just as concerned about multiple time servers (e.g. > docsDevServerTimeAddress) as about multiple SYSLOG servers (e.g. > docsDevEvSyslogAddress). > > RFC 2669 (the original Cable Device MIB) was published August 1999, > and > RFC 2132 was published March 1997. So the MIB module restrictions for > a > single time server and SYSLOG server has been around for over six > years, > without an objection until now. Is this a real problem, or a > theoretical > problem? > > I can think of three reactions to this objection: (1) no changes, (2) > clarify the relevant MIB objects so that they report the *first* DHCP > option value if the DHCP option is multi-valued, or (3) change the MIB > module to add tables to capture (multiple) Time Servers and/or SYSLOG > servers. > > -- Rich > > -----Original Message----- > From: [email protected] [mailto:[email protected]] On Behalf > Of Wijnen, Bert (Bert) > Sent: Wednesday, December 21, 2005 12:39 PM > To: Ipcdn (E-mail) > Cc: Brian Hedstrom > Subject: RE: [ipcdn] Last Call: 'Cable Device Management Information > Base for Data-Over-Cable Service Interface Specification Compliant > CableModem s and Cable Modem Termination Systems' to Proposed Standard > > > Authors/Editors and WG, > > A comment on IETF Last Call for your consideration and > possible action. > > Not sure that Brian is subscribed to IPCDN list, so pls DO > copy him on any responses. > > Bert > > > > > -----Original Message----- > > From: [email protected] [mailto:[email protected]]On Behalf > Of > > > Brian Hedstrom > > Sent: Monday, December 19, 2005 20:23 > > To: [email protected] > > Subject: RE: [ipcdn] Last Call: 'Cable Device Management Information > > Base for Data-Over-Cable Service Interface Specification Compliant > > Cable Modems and Cable Modem Termination Systems' to Proposed > Standard > > > > > > IETF: > > > > Regarding this object: > > docsDevEvSyslogAddress OBJECT-TYPE > > SYNTAX InetAddress > > MAX-ACCESS read-write > > STATUS current > > DESCRIPTION > > "The internet address of the Syslog server. If the > > address of the server is set to any of the zero > length > > string, the 0.0.0.0 V4 address or the 0: V6 address, > > syslog transmission is inhibited. By default at > agent > > boot, this object returns the zero length string." > > ::= { docsDevEvent 10 } > > > > This object does not support provisioning multiple Syslog > > servers within > > the CM. However, DHCP Option 7 (log server option, as > > defined in DOCSIS > > 2.0 Radio Frequency Interface Specification > > [CM-SP-RFIv2.0-I09-050812]) > > does allow for specification of multiple log server addresses so > > although they cannot all be instrumented it is possible to provide > the > > CM with multiple log server addresses. It seems reasonable to > > extend the > > docsDevEvSyslogAddress object to support multiple IP > > addresses but I do > > not know if this was already addressed and rejected during the > > development of this Internet draft. The change could potentially > mean > > deprecating docsDevEvSyslogAddress and replacing it with a > > table object > > which supports Row Creation or statically creates n-number of rows. > A > > similar approach was taken in CableHome with the definition of > > cabhCdpWanDnsServerTable (CABH-CDP-MIB). The structure of > > cabhCdpWanDnsServerTable identifies an order of preference (primary, > > secondary, or tertiary) for each entry. See below: > > cabhCdpWanDnsServerTable OBJECT-TYPE > > SYNTAX SEQUENCE OF CabhCdpWanDnsServerEntry > > MAX-ACCESS not-accessible > > STATUS current > > DESCRIPTION > > "This table contains the IP addresses of cable network > and > > Internet DNS servers, in the order of preference in > which > > the PS's CNP will query them, when it cannot resolve a > DNS > > query using local information. Entries in this table are > > updated with the information contained in DHCP Option 6, > > received during both the WAN-Man and WAN-Data IP > > acquisition processes." > > ::= { cabhCdpAddr 3 } > > > > cabhCdpWanDnsServerEntry OBJECT-TYPE > > SYNTAX CabhCdpWanDnsServerEntry > > MAX-ACCESS not-accessible > > STATUS current > > DESCRIPTION > > "List of cable network and Internet DNS servers." > > INDEX { cabhCdpWanDnsServerOrder } > > ::= { cabhCdpWanDnsServerTable 1 } > > > > CabhCdpWanDnsServerEntry ::= SEQUENCE { > > cabhCdpWanDnsServerOrder INTEGER, > > cabhCdpWanDnsServerIpType InetAddressType, > > cabhCdpWanDnsServerIp InetAddress > > } > > > > cabhCdpWanDnsServerOrder OBJECT-TYPE > > SYNTAX INTEGER { > > primary(1), > > secondary(2), > > tertiary(3) > > } > > MAX-ACCESS not-accessible > > STATUS current > > DESCRIPTION > > "The order of preference for cable network and > > Internet DNS > > servers, as listed in DHCP option 6 (Domain Server). Any > > time the CDC receives valid IP address information > within > > DHCP Option 6, as part of lease acquisition or renewal > of > > a WAN-Man or WAN-Data IP, it must update this > information > > into this table. As entries in DHCP Option 6 are > > listed in > > order of preference, the highest priority entry in DHCP > > Option 6 must correspond to the row with a > > cabhCdpWanDnsServerOrder with a value of 1. If DHCP > > Option 6 contains 1 valid IP address, the PS MUST update > > the row with a cabhCdpWanDnsServerOrder value of 1 and > > MUST NOT modify rows with > > cabhCdpWanDnsServerOrder values of 2 & 3 > > (if they exist). If DHCP Option 6 contains 2 valid > > IP addresses, the PS MUST update the rows with > > cabhCdpWanDnsServerOrder values of 1 and 2 > > and MUST NOT modify the row with > cabhCdpWanDnsServerOrder > > value of 3 (if it exists). If DHCP Option 6 contains 3 > > valid IP addresses, the PS MUST update rows with > > cabhCdpWanDnsServerOrder values of 1, 2, and 3. > > Any DNS server information included in DHCP Option 6 > > beyond primary, secondary and tertiary will not be > > represented in this table." > > ::= { cabhCdpWanDnsServerEntry 1 } > > > > cabhCdpWanDnsServerIpType OBJECT-TYPE > > SYNTAX InetAddressType > > MAX-ACCESS read-only > > STATUS current > > DESCRIPTION > > "This parameter indicates the IP address type of a > > WAN DNS server." > > DEFVAL { ipv4 } > > ::= { cabhCdpWanDnsServerEntry 2 } > > > > cabhCdpWanDnsServerIp OBJECT-TYPE > > SYNTAX InetAddress > > MAX-ACCESS read-only > > STATUS current > > DESCRIPTION > > "This parameter indicates the IP address of a WAN DNS > > server. The type of this address is specified by > > cabhCdpWanDnsServerIpType." > > ::= { cabhCdpWanDnsServerEntry 3 } > > > > Thanks, > > Brian Hedstrom > > Senior Engineer, Operation Support Systems > > Broadband Access > > CableLabs, Inc. > > 858 Coal Creek Circle > > Louisville, CO 80027 > > Direct: 303.661.3829 > > eFax: 303.664.8120 > > [email protected] > > > > > > -----Original Message----- > > From: The IESG [mailto:[email protected]] > > Sent: Wednesday, December 14, 2005 9:44 AM > > To: IETF-Announce > > Cc: [email protected] > > Subject: [ipcdn] Last Call: 'Cable Device Management Information > Base > > for Data-Over-Cable Service Interface Specification Compliant Cable > > Modems and Cable Modem Termination Systems' to Proposed Standard > > > > The IESG has received a request from the IP over Cable Data Network > WG > > > to consider the following document: > > > > - 'Cable Device Management Information Base for > > Data-Over-Cable Service > > Interface Specification Compliant Cable Modems and Cable Modem > > Termination > > Systems ' > > <draft-ietf-ipcdn-device-mibv2-10.txt> as a Proposed Standard > > > > This document obsoletes RFC2669. > > > > The IESG plans to make a decision in the next few weeks, and > solicits > > final comments on this action. Please send any comments to the > > [email protected] or [email protected] mailing lists by 2005-12-28. > > > > The file can be obtained via > > http://www.ietf.org/internet-drafts/draft-ietf-ipcdn-device-mi > bv2-10.txt > > > _______________________________________________ > IPCDN mailing list > [email protected] > https://www1.ietf.org/mailman/listinfo/ipcdn > > > _______________________________________________ > IPCDN mailing list > [email protected] > https://www1.ietf.org/mailman/listinfo/ipcdn > > _______________________________________________ > IPCDN mailing list > [email protected] > https://www1.ietf.org/mailman/listinfo/ipcdn