IPP> NOT - ISSUE 1 - can event reporting be dropped if device is too b usy?

IPP> NOT - ISSUE 1 - can event reporting be dropped if device is too b usy?

IPP> NOT - ISSUE 1 - can event reporting be dropped if device is too b usy?

don at lexmark.com don at lexmark.com
Fri May 21 17:36:07 EDT 1999


I think a agree with Ira on this.  If the subscription method is unreliable than
dropped events from whatever cause should be expected.  If the subscript method
is reliable then they must not be dropped; rather they might be delivered
"late."

Don




imcdonal%sdsp.mc.xerox.com at interlock.lexmark.com on 05/19/99 12:41:09 PM

To:   hastings%cp10.es.xerox.com at interlock.lexmark.com,
      ipp%pwg.org at interlock.lexmark.com
cc:    (bcc: Don Wright/Lex/Lexmark)
Subject:  Re:  IPP> NOT - ISSUE 1 - can event reporting be dropped if device is
      too b usy?




Hi Tom,

Since most event reports are delivered through inherently
unreliable (or anyway unconfirmed) channels, I'd say ANY
device MAY suspend event reporting when busy.

Rather than have clients indicate that their events are
'droppable' (low-priority), I'd rather that the default
is that all event subscribtions are for 'unconfirmed'
event delivery service (e.g., SNMP Trap-PDU or SMTP
'mailto:' without MDN or DSN feedback).  Then a client
can explicitly subscribe (if the device offers it) to
the 'confirmed' event delivery service (e.g., a
response from a SNMP management station/ trap
destination via a 'latch' object in the device's MIB
or use of MDN/DSN in the SMTP 'mailto:' case).

Cheers,
- Ira McDonald
  High North Inc








More information about the Ipp mailing list