[IPFIX] RFC 8158 applicability and breadth

"Wayne Tackabury" <[email protected]> Mon, 15 Jan 2018 22:35:12 +0000
Newsgroups gmane.ietf.ipfix
Message-ID <OFF41D978F.7EA008FF-ON00258216.007B29FE-00258216.007C1282@notes.na.collabserv.com>
--===============3086852804016965674==
Content-Transfer-Encoding: quoted-printable
Content-Type: text/html; charset=UTF-8

<div class=3D"socmaildefaultfont" dir=3D"ltr" style=3D"font-family:Arial, H=
elvetica, sans-serif;font-size:10.5pt" ><div dir=3D"ltr" >Hi all:</div>
<div dir=3D"ltr" >&nbsp;</div>
<div dir=3D"ltr" >I have two questions, and will separate them since they'r=
e distinct, IPFIX-wise.</div>
<div dir=3D"ltr" >&nbsp;</div>
<div dir=3D"ltr" >I'm noting the presence of RFC 8158, have briefly gone ov=
er it.&nbsp; It addresses IEs for "logging NAT events".</div>
<div dir=3D"ltr" >&nbsp;</div>
<div dir=3D"ltr" >My question is this: in the formulation of these entities=
, was the intent to have this be strictly the reporting domain for "classic=
" NAT implementations, with configured pools and ports and static vs. dynam=
ic mappings and all of that?&nbsp;</div>
<div dir=3D"ltr" >&nbsp;</div>
<div dir=3D"ltr" >Or, was there a considered "mission creep" :) to allow th=
ese to be used for reporting on things like load balancers, firewall domain=
 address mapping, and the like, to report what their "pre-translated" addre=
ss is on the source of a flow?&nbsp; Keep in mind, the mapping for such a t=
hing has nothing to do with the exporter itself (the exporter may just be d=
etecting this from network traffic), so it would be quite wrong to use a so=
urce vs. exporter address.</div>
<div dir=3D"ltr" >&nbsp;</div>
<div dir=3D"ltr" >If using the RFC 8158 fields for this just seems....wrong=
 (I mean, it is translation....of an address), is there another (probably s=
impler) set of fields for this that come to mind?&nbsp; To be perfectly tra=
nsparent here, if this was to be used for supporting the semantics of the "=
Forwarded: " HTTP header as in RFC 7239, this would in fact have to be a li=
st of addresses, but I'm just wondering if there was an IPFIX reporting pra=
ctice for this using IANA elements even if the forwarded source was a scala=
r address.</div>
<div dir=3D"ltr" >&nbsp;</div>
<div dir=3D"ltr" >Thanks!</div>
<div dir=3D"ltr" >Wayne</div>
<div dir=3D"ltr" >&nbsp;</div></div><BR>


--===============3086852804016965674==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
IPFIX mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/ipfix

--===============3086852804016965674==--