Hi Venu,

Thanks for reading the draft and posting queries.

See inline...

From: Pce [mailto:[email protected]] On Behalf Of Venugopal Reddy K
Sent: 26 October 2015 14:39
To: [email protected]; pce
Subject: [Pce] Few queries regarding 
draft-dhody-pce-stateful-pce-auto-bandwidth-06 draft

Dear Authors,

I have few queries regarding draft-dhody-pce-stateful-pce-auto-bandwidth-06.


1.       We have Overflow-Count, Overflow-Threshold and Underflow-Count, 
Underflow-Threshold to expedite the process of PCC adjusting the bandwidth 
instead of waiting for Adjustment-Interval when an abrupt change in bandwidth 
usage occurs. Similarly, we have Report-Flow-Threshold value for a PCC to 
report for PCE adjusting bandwidth model instead of waiting for 
Report-Interval. Can we know if there is rationale behind having different 
method for PCC controlling bandwidth adjustment and PCE controlling bandwidth 
adjustment for this case? Why not have the similar mechanism for it ?


[Dhruv]: Are you asking why we don't have a "Report-Flow-Count" (similar to 
Underflow and overflow-count)? Hmm, I guess "Report-Flow-Count" can be added, 
it is a good suggestion.
This would indicate how many times consecutively, the percentage or absolute 
difference between the current bandwidth sample and the current bandwidth 
reservation should be greater than or equal to the report-flow-threshold value, 
for all the bandwidth samples collected so far to be reported to the PCE 
immediately.



2.       Once auto-bandwidth adjustment is enabled for an LSP, Couldn't find 
any way to disable auto adjustment feature for an LSP. It would be good to have 
a means to disable adjustment without bringing the LSP down.  Also, changing 
auto-bandwidth adjustment mode from PCC to PCE or vice versa without LSP down.

[Dhruv]: I think this could be useful, will discuss with my co-authors and add 
mechanism for this in future revision.

Regards,
Dhruv


Could you please clarify.

Thanks :)

Regards,
Venu
_______________________________________________
Pce mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/pce

Reply via email to