Re: [atlas]Feature request (fo consideration): broadcast rate measurement?
Hi As I see it the atlas project is about getting a global view. So in my book the last thing we want is stats about the local LAN segment the probe is connected to. Further more for privacy and security reasons the probes should in my mind gather as little information about the directly attached network as possible. /Jesper "Wilfried Woeber, UniVie/ACOnet" <Woeber@CC.UniVie.ac.at> wrote:
Xmas WishList 2011 :-)
Looking at the yellow LED of the probes, I presume that their Ethernet IF operates in promiscuous mode. Even if not...
I'd like to see some (easily configurable) measurement available, at least the rate of broadcast-packets on the probe's net.
On a more elaborate basis, maybe according to simple paramters like a protocol/port combination.
Would this make sense to implement and be of interest to others?
Thanks for consideration, Wilfried.
PS: yes, I do know about Netflow, WireShark and similar, but the beauty of the ATLAS Probe Approach is imho in the ready-nade UI :-)
On 09.04 12:42, Jesper Skriver wrote:
Hi
As I see it the atlas project is about getting a global view. So in my book the last thing we want is stats about the local LAN segment the probe is connected to. Further more for privacy and security reasons the probes should in my mind gather as little information about the directly attached network as possible.
/Jesper
I agree. Whhile this may be interesting it is the first step on a slippery slope towards passive measurements which would endagner confidence in the "we are not listening to your traffic" message. Daniel
Good points, thanks! Wilfried Daniel Karrenberg wrote:
On 09.04 12:42, Jesper Skriver wrote:
Hi
As I see it the atlas project is about getting a global view. So in my book the last thing we want is stats about the local LAN segment the probe is connected to. Further more for privacy and security reasons the probes should in my mind gather as little information about the directly attached network as possible.
/Jesper
I agree. Whhile this may be interesting it is the first step on a slippery slope towards passive measurements which would endagner confidence in the "we are not listening to your traffic" message.
Daniel
participants (3)
-
Daniel Karrenberg -
Jesper Skriver -
Wilfried Woeber, UniVie/ACOnet