Logarithmic scale

Sorry that is not possible until now.

Since this more than a year back and the answer was "... until now." I wanted to ask if there is a possibility for this now (or planned)?

Thank you for the answer.

For me the graphics would be a very good source to give me an idea of what the system load looks like but since they only support linear scale their use is limited. To give a good example where it would be necessary to be able to see "something": A HDD which stores a 20 GB database and has an average read rate of 5 MB/s. Each morning there will be a full backup which runs at 300 MB/s. Now when I look at the graph I see only a 300 MB/s spike but the rest is basically around 0 since the graphics are scaled to allow the maximum to be visible. Of course I could set a spike filter and exclude the backup spike but that would show wrong information then.

There are many other examples where short "valid" spikes cause the more important parts of a graphic show basically as 0 and a logarithmic scale should fix that.

I would also like logarithmic scales for graphs. This makes it easier see what's going on when there are large variations in data values.

It's the only way get a good idea of traffic/bandwidth usage to be honest. I hope the dev team would re-consider implementing logarithmic interpretation of data.

I can assure you if there was a logarithmic feature, a lot of people would use it. Just because it's not asked for a lot doesn't mean it's not a feature people would use :)

Me too, I think the only way to have a good point of view for data that can have such a large variance is the logarithmic scale. Then another request for this feature!

This would be extremely beneficial, please add this!

I am very interested too to have log scales because a tiny value is readable as a huge one. In some graphics is very usefull. Thanks to add it.

You now have several requests for this- including mine. 8^) It's pretty useless for some things without this.

+1 for logarithm scale of y-axis

One more Sample. I do synthetic HTTP-Requests (Custom XML Sensor) to measure application latency and i'm logging Min/Max/Avg. So min is about 10-20ms, Avg about 15-45ms and MAX is betweeen 20ms and even 2000ms. So i have to hide the max graph so see the details of the Min(/Avg values.

A logarithm scaling would be great.

Do we need to open a new issue so that this gets attention?

This would be amazing! I came here looking for how to switch the axis to logarithmic and I'm surprised this isn't already a feature. With interfaces that jump to max bandwidth usage during backup operations, it's really hard to see fluctuations when backups aren't taking place. It just goes from huge plateau to almost zero.

The last response to this from Paessler, was from back in 2014. Does anyone know if this will ever be implemented?

Dear claed,

while nothing is not ruled out, please don't expect an implementation for a logarithmic scale. When we get features requested, we look at the overall demand and prioritize accordingly.

Please implement this.

Dear gfox,

you can create an official feature request, the procedure is described here. Please allow me to say that nonlinear graph axes are not a top priority because there are a lot of other features in higher demand, some of those are listed on the roadmap.

I have created an official feature request. It should appear within 24 hours. If those who have participated in this thread who haven't given up on PRTG would up-vote it when it appears, that may help it to happen.