testing bandwidth restrictions

[eve3091]

I understand the restriction is set in MB. Am I wrong in thinking that I should be able to enter nos like eg. 0.1 because I did that and saw only a very small change in the bandwidth used, no matter what no I entered there. As an example, here the actual 5 min avg dvg graph while I was testing.

You can see 3 spikes starting at ~1200. Each of them shows 1 test. I ran the tests just long enough for them to show up in the graph which turned out to be ~5 to 10 mins each which is why they show up as spikes only on this graph. On the first spike, the bandwidth was unlimited, on the 2nd one I had set it to 0.1 MBps and on the 3rd one I had set it to 0.001 MBps. I have been working on other things since ~0800 as can be seen on the graph but far from enough to influence the numbers much. As far as I can tell, the settings do influence the bandwidth used a little bit but eg. the 0.001 MBps is far from being 1/100 of 0.1 MBps.

[superflexible]

Yes you can specify numbers like 0.1.

Your chart does not show the average bandwidths, it shows the maximum ones. I think it's the wrong chart for this topic. Bandwidth limitation will limit the average transfer over time, but individual chunks of the buffer sizes used may still be transferred at high speed, followed by a pause.

[eve3091]

Ok, I' reproduce this and try and take some screenshots of our monitoring software that gives a better overview on the actual data transferred and how it was transferred. We have a backup system in place for our newer machines that works exactly as you describe and there the data transfer graph looks like the pulse on a life support system

Oh...and btw. thank you!!! for answering on a weekend

[eve3091]

Ok, I did reproduce this with a better graph and it did show that setting the restriction to 0.001MB transferred the data at a rate of ~2 to 5 Mbps with the occasional spike. Going down even further to 0.0001 truncated the number but it also had the effect of lowering the avg. transfer rate to ~0.8Mbps.

In the end, I installed a proper ftp server on the server and set the rate limit there so all is well now

If you are interested let me know. I still have the graphs and as inaccurate as they are for this type of problem they do show that the restriction built into SuperFlexibleFileSynchronizer isn't enforced as closely as somebody who pays bandwidth on a 95%ile would like If you'd need to test let me know; I'd be happy to help.

[superflexible]

Hi,

thanks for the info! I will look into it.