bandwidth chart is inaccurate

[Wayne Workman]

So I’ve figured out that the bandwidth chart is off in relation to the delay set.

I’ve got it set to 3 second interval, and it’s off by x3. So sites that are 100Mbps are maintaining 300Mbps during imaging on the bandwidth graph. Sites that are 1Gbps are showing 3Gbps steady on the graph when it’s at capacity.

When I had it set to 6 seconds, it was inaccurate by x6.

[ITSolutions]

@Wayne-Workman I have to ask the question that you ask and we all ask, What version are you on? And did it just start in a recent update?

[Wayne Workman]

@ITSolutions Version 8157, I’ve not updated in a while, I was just reporting the bug.

[Tom Elliott]

You mean like this one?

[Tom Elliott]

https://forums.fogproject.org/topic/7921/bandwidth-graph-8215-scale-is-wrong

[Wayne Workman]

@Tom-Elliott No, not the same thing.

In this thread, the chart was reporting x4 and x6 and x10 times more than what was actual, whereas in that new thread you linked to, the chart is showing x10 times less than actual.

[Tom Elliott]

Well what is the basis of “inaccurate” then?

[Wayne Workman]

@Tom-Elliott lol the scale being wrong.

[Tom Elliott]

But how is it wrong?

How do you know it’s displaying incorrectly?

Is this one fixed or not? Is the other incorrect or not? How do you know the scale is wrong?

[Wayne Workman]

@Tom-Elliott I look in task management and see how many downloads are going for that site and their speeds, and I can estimate that particular building’s 100Mbps network is going at capacity based on my own experiences with what partclone displays at home, my building, and other places.

[Tom Elliott]

@Wayne-Workman Except partclone is working with compressed data. The data sent will be slowing than the data received. 10 GB a min is already more than a gig network can handle, but this isn’t because the network is going faster.

We need a way to benchmark actual bandwidth speeds on a node/server with what is being displayed on the bandwidth graph.

[Wayne Workman]

@Tom-Elliott Tom, I know this.

Imaging is happening and completing in 16 minutes at that site on a 100Mbps network, the image is 19 ish GB in size. The bandwidth chart is wrong.

[Tom Elliott]

@Wayne-Workman How is it wrong? We need real data. I can’t seem to see where the inaccuracy lies.

[Wayne Workman]

After logging out and logging in, the scale on the bandwidth chart is fixed. I did enable another storage node that was down, two nodes are still disabled because they are still down.

[Tom Elliott]

Is it safe to solve this thread then? The chart adjusts to all nodes bandwidth btw. This means if one node has only used 10Mbps and another has used 100Mbps, the 10Mbps will look like it’s much less as it will scale to the largest node.

[Wayne Workman]

@Tom-Elliott I guess. The graph appears to be exactly correct at the moment.