Re: Re: sFlow Datagram Extensibility

From: Marc Lavine (mlavine@foundrynet.com)
Date: 09/24/02

  • Next message: Marc Lavine: "Distributed agent implementation"

    > > I view the information about a discard having occurred on
    > > input or output as
    > > pretty architecture-independent, although the interpretation of that
    > > information would be architecture-dependent. I guess we can
    > > always add a
    > > vendor extension if we need this information, though.
    >
    > Adding this information as a vendor extension, or even as a standard sFlow
    > data record would be possible without having to change SFLOW-DATAGRAM
    > version. A detailed discard record could even have a reference to the
    > particular ACL entry that caused the discard.

    Yes, I agree. That flexibility is one of the big benefits of the version 5
    changes.

    In the flow_sample, we've got:

       interface input; /* Interface packet was received on. */
       interface output; /* Interface packet was sent on. */

    And the "interface" typedef defines formats for these values, including the
    discard reason codes. Previously, the discard codes were only present in
    the output field. I think we need to define their applicability to these
    two fields. Perhaps we should we indicate that while the "single interface"
    values may appear in either field, the discard codes and the multiple
    interface codes should only be used in the output field?

    Marc



    This archive was generated by hypermail 2.1.4 : 09/24/02 PDT