Re: Comments on draft-ietf-msgtrk-mtqp-03.txt
Tony Hansen <[email protected]> Sat, 04 Aug 2001 14:07:05 -0400
| Newsgroups | gmane.ietf.msgtrk |
|---|---|
| Message-ID | <[email protected]> |
See comments below. Tony Gregory Neil Shapiro wrote: > > 2. Basic Operation > > The second paragraph describes how to determine tracking host to contact. > How about a host like this: > > gshapiro.net IN SRV mtqp tracking.gshapiro.net > IN MX 10 smtp.gshapiro.net > IN MX 20 smtp.sendmail.com > > In other words, a dedicated tracking server for gshapiro.net and two MX > records. When an SMTP message is sent, smtp.gshapiro.net happens to be > unavailable so the message goes to smtp.sendmail.com. If an MTQP client > then tries to track the message, it will use tracking.gshapiro.net (due to > the SRV record). However, tracking.gshapiro.net doesn't know about the > message yet (still in smtp.sendmail.com's queue). The lookup fails. > > Also, it's not clear from the paragraph that if SRV records do not exist, > do you only use the first MX returned or all of them in MX preference > order? > > If I may suggest the following change: > > When an MTQP client wishes to make use of the message tracking > service, it establishes a TCP connection with the server host. To > find the server host, the MTQP client first does an MX lookup for the > server host using DNS MX records, as specified in [RFC-DNS] and > revised by [RFC-HOSTS]. If no MX records are found, the MTQP client > then does an A record lookup for the server host. If MX records are > found, for each MX record (in MX preference order -- lowest to > highest), an SRV lookup on the MX host name using DNS SRV records, > with a service name of "mtqp". (See the "Usage rules" section in > [RFC-SRV] for details.) If a SRV host is found, the MTQP client uses > that host for tracking. Otherwise, it uses the MX host name for > tracking. Finally, the host either the SRV host (if found) or the MX > host is contacted using MTQP. This process is repeated for each of > the MX records until an MTQP lookup succeeds (i.e., tracking status is > returned). > > With this new text, the DNS records might look something like: > > gshapiro.net IN MX 10 smtp.gshapiro.net > IN MX 20 smtp.sendmail.com > > smtp.gshapiro.net IN A 10.254.153.10 > > smtp.sendmail.com IN A 192.168.10.43 > IN SRV mtqp tracking.sendmail.com > > tracking.sendmail.com IN A 192.168.10.45 > > In which case, a message sent to [email protected] would be tracked by > first contacting smtp.gshapiro.net with MTQP and if no tracking information > is available at that host (or it can not be contacted), then continue by > contacting tracking.sendmail.com with MTQP. > > An example, such as the one I have given should be added to the draft. I don't know how common this usage of MX records is, but it is valid. I'm okay with this change. Anyone else? > --- > > 3. Initialization and Option Response > > Example 5 should probably use "VND." as Option2 and Option3 are not valid > (only STARTTLS is). A suggested replacement: > > Example #5 (options available): > S: +OK+/MTQP MTQP server ready > S: starttls > S: Vnd.vendor.Option2 with parameters > S: Vnd.vendor.Option3 with a very long > S: list of parameters > S: . sure > --- > > 10. IANA Considerations > > Vendor-specific options MUST be registered with IANA? I thought the whole > point of vendor options was like that of 'X' ESMTP extensions: > > RFC 2821 Simple Mail Transfer Protocol April 2001 > 2.2.2 Definition and Registration of Extensions > > In addition, any EHLO keyword value starting with an upper or lower > case "X" refers to a local SMTP service extension used exclusively > through bilateral agreement. Keywords beginning with "X" MUST NOT be > used in a registered service extension. Conversely, keyword values > presented in the EHLO response that do not begin with "X" MUST > correspond to a standard, standards-track, or IESG-approved > experimental SMTP service extension registered with IANA. A > conforming server MUST NOT offer non-"X"-prefixed keyword values that > are not described in a registered extension. > > If that is the case, then they should not be registered with IANA. Perhaps > we want to steal some of the above text for this document. I meant to say that the vendor name must be registered, as is done with MIME content types. Another option is to go the java route and use a reversed domain name: vnd.com.example.option1 This avoids the IANA registration issue entirely, under the assumption that only the owners of a domain would use their domain. > --- > > 12. Protocol Syntax > > Move opt-text up as it is used by comment-command first (before > temp-response). > > The response-info ABNF is incorrect. The parens are not balanced. ok > --- > > Spelling fixes: > > @@ -197 +197 @@ > -parseable, case-insensitive response information giving more data about > +parsable, case-insensitive response information giving more data about > @@ -249 +249 @@ > -Which option is picked is an adminstrative decision and is not further > +Which option is picked is an administrative decision and is not further > @@ -667 +667 @@ > - System port number XXXX - TBA by IANA > + System port number XXXX - TBD by IANA > @@ -693 +693 @@ > -begin with "vnd." MUST be registered with IANA on a Firt Come First > +begin with "vnd." MUST be registered with IANA on a First Come First > @@ -721 +721 @@ > - Both the STMP client and server must check the result of the TLS > + Both the SMTP client and server must check the result of the TLS > @@ -725 +725 @@ > -was achieved is made locally, is implementation-dependant, and is beyond > +was achieved is made locally, is implementation-dependent, and is beyond > @@ -825 +825 @@ > -13. Acknowledgements > +13. Acknowledgments > @@ -925 +925 @@ > -assist in its implmentation may be prepared, copied, published and dis- > +assist in its implementation may be prepared, copied, published and dis- > @@ -930 +930 @@ > -references to the Internet Society or other Internet organisations, > +references to the Internet Society or other Internet organizations, ok