Re: RAQMON Trap PDU size and report transmission time
Juergen Schoenwaelder <[email protected]>
| Newsgroups | gmane.ietf.rmonmib |
|---|---|
| Message-ID | <20050118190304.GB3016@james> |
On Mon, Jan 17, 2005 at 07:48:35PM -0800, Randy Presuhn wrote: > > This is per RDS. A RDS can generate at most 1 RAQMON Trap PDU > > every 2 minutes, regardless of the number of flows (or report > > fragments). This rate limit does not apply to Inform PDUs, > > just SNMP Trap PDUs over UDP. > ... > > In light of this constraint, is there any case where mapping to a trap, > rather than an inform, would ever make sense? Should we just > say that, operationally, configuring a system to send these as traps > is NOT RECOMMENDED, since the congestion avoidance could > cause much information to be lost? It seems the authors assume "usage of Inform PDUs" implies "stop-and-wait" behaviour and they seem to assume that "stop-and-wait" with N end devices (N >> 1) does not cause congestion in the network. /js -- Juergen Schoenwaelder International University Bremen <http://www.eecs.iu-bremen.de/> P.O. Box 750 561, 28725 Bremen, Germany