[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[PSAMP] a few comments on the v4 framework draft
Hi, here are a few comment on the last version of the PSAMP framework draft.
Regards,
Maurizio
Sec.2, page 5 – “Export packets: one or packet
reports…”
Should be “… one or more packet reports….”
Sec. 4.1, page 9. Before the definition of “Hash Selection Range”,
add the definition of “Hash Value: the result of the action of the hash
function on the hash domain”. Then change the definition of Hash
Selection Range to “….the packet is selected if the Hash Value
yields a result in the Hash Selection Range”. Finally, copy also these
modifications in the terminology section of the Sampling Technologies draft
(where the term “Hash value” is frequently used…)
Sec. 4.1, page 9. “Population size: the number of packets in a subset of
a packet stream”.
Should be “…the number of packets in a packet stream”
Sec.4.1, page 9. “Target selection Frequency:…..Depending on the
selector, the target selection frequency may be count-based or time
based”
Suggestion: remove the sentence starting with “Depending…”,
otherwise it’s not coherent with the definition of the “Attained
selection frequency”, which is only count based.
Section 5.1 “Mandatory content of Packet Reports”
The title is a little bit misleading, and in fact it is not mandatory to insert
(ii) in the packet reports: what is mandatory is that there MUST be the
capability of doing so, as the text clearly says.
NOTE: the author of the protocol draft was actually probably mislead and wrote
in section 7.1 that the packet report MUST also contain (ii), which is not
true….
Suggestion: change the title or split 5.1 in two.
Section 5.5 (Report Timeliness), third paragraph, “A dispatch delay of 1
second in these situations eliminates the need for timestamping by synchronized
clocks at observation points devices, or for the observation points and
collector to maintain bi-directional communication in order to track clock
offsets”…
This is not true for all the applications (e.g. One Way Delay estimation).
Suggestion: either remove the text or specify that this holds true only for
certain applications.
--
--
--
to unsubscribe send a message to psamp-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.
archive: <http://ops.ietf.org/lists/psamp/>