light blue peak on graph

Nikolay Shopik

21 Oct 2011 21 Oct '11

6:51 p.m.

So I got one device polled by cacti and observium. And see difference in peaks (outbound as example). What I see is light blue color peaks are bigger than show in cacti but regular blue outline in observium graph match cacti peak. Is light blue peak is interpolation of data, how its caclucated or where it comes from?

Show replies by date

adama.armstrong＠gmail.com

21 Oct 21 Oct

6:55 p.m.

Light blue is peak of peak. Cacti might be peak of ave

-- Sent from my Android phone with K-9 Mail. Please excuse my brevity. Nikolay Shopik shopik@inblock.ru wrote: So I got one device polled by cacti and observium. And see difference in peaks (outbound as example). What I see is light blue color peaks are bigger than show in cacti but regular blue outline in observium graph match cacti peak. Is light blue peak is interpolation of data, how its caclucated or where it comes from? _____________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium

adama.armstrong＠gmail.com

6:58 p.m.

Specifically, its the maximum value of the MAX rra. Cacti probably uses maximum of average, or average of maximum.

Look at the graphtype code to see!

Nikolay Shopik

7:06 p.m.

So in that case observium report more correct peaks, with cost more zoomout on graph. It will be nice have option in config to change from max to ave. And personally I think max looks good on fullscreen graph, but on any smaller scale it start to ruin graphs.

But thanks for pointer where to look at.

On 21.10.2011, at 20:58, "adama.armstrong@gmail.com" adama@memetic.org wrote:

...

Specifically, its the maximum value of the MAX rra. Cacti probably uses maximum of average, or average of maximum.

Look at the graphtype code to see!

Sent from my Android phone with K-9 Mail. Please excuse my brevity.

Nikolay Shopik shopik@inblock.ru wrote: So I got one device polled by cacti and observium. And see difference in peaks (outbound as example). What I see is light blue color peaks are bigger than show in cacti but regular blue outline in observium graph match cacti peak. Is light blue peak is interpolation of data, how its caclucated or where it comes from?

observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium _______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium

adama.armstrong＠gmail.com

7:24 p.m.

I think allowing that to be changed would be very confusing. We actually need to standardise more!

I agree that it makes the graphs scale strangely.

I also want to make those graphs prettier. They don't match with the multibits grapes at all!

-- Sent from my Android phone with K-9 Mail. Please excuse my brevity. Nikolay Shopik shopik@inblock.ru wrote: So in that case observium report more correct peaks, with cost more zoomout on graph. It will be nice have option in config to change from max to ave. And personally I think max looks good on fullscreen graph, but on any smaller scale it start to ruin graphs. But thanks for pointer where to look at. On 21.10.2011, at 20:58, "adama.armstrong@gmail.com" adama@memetic.org wrote: Specifically, its the maximum value of the MAX rra. Cacti probably uses maximum of average, or average of maximum. Look at the graphtype code to see! -- Sent from my Android phone with K-9 Mail. Please excuse my brevity. Nikolay Shopik shopik@inblock.ru wrote: So I got one device polled by cacti and observium. And see difference in peaks (outbound as example). What I see is light blue color peaks are bigger than show in cacti but regular blue outline in observium graph match cacti peak. Is light blue peak is interpolation of data, how its caclucated or where it comes from? _____________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium _______________________________________________ observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium

4857

Age (days ago)

4857

Last active (days ago)

List overview

Download

4 comments

2 participants

tags (0)

participants (2)

adama.armstrong＠gmail.com
Nikolay Shopik