Canceled Downloads

Hey, is it normal that successful download rate can be down to like 60% without any network or IO bottleneck on my server side?

I have 2 node running in parallel on my server, but they are on seperate HDD, IO wait is in good levels, but I just started to monitor my nodes and I see the following table. Both node is about the same size and same age.

Could it be that just “my” user is closer to someone else with the same data? Its just interesting to see that the succesful download rate is about 60% for my node in the last day.

They only need 29 of 39 downloads, therefore the average sucess rate should be like 74%. So 60% seems to be within normal fluctuation.

1 Like

This might just as well be a single customer who happens to be far from you.

1 Like

In short - all customers’ behavior is normal. Likely your node is not fast enough for this (or these) customer(s).

hello, How do you monotoryour nodes ? i’m interested your method.
can you explain me how (logs, grab data, etc …)

thanks for reply
philippe (France)

Thanks for the reply (and @ Toyoo)

It was just interesting to see this big fluctuation while the downloaded amount was quite the same.

Hey, I followed the official documentation about it.

So it is through the debug port, Prometheus and Graphana

Yeah, Storj is small enough that a single bigger customer’s activity has a big impact on general traffic patterns observed by a single node. I suspect Storj needs to grow 2-3 orders of magnitude to average them out.

1 Like