So, I’m not exactly sure what has happened here. I started two new nodes on one of my servers and my bandwidth spiked through the roof. I’m assuming its because I changed the net.core.rmem_max value to the recommended value of 2500000? Would this cause an increase in bandwidth?
What is the default value for ubuntu?
In the mean time I’ve cut the value in half to 1250000 to see if this helps.
It’s likely related to the customers activity, not your changes. However, if this initial change helped to fix QUIC, your node just win more customers.
I do not see a reason why do you need to make it less popular?
A typical storagenode uses a few mbit/s of internet bandwidth (average, actual usage is a little bursty/spikey but seldom exceeds 20mbit/s for more than a few seconds)
Your graph shows much higher usage. I can only conclude the traffic is something unrelated to the storagenodes you run.
After further investigation, it was an internal transfer that just happened to start when I started the two nodes causing the spike, not node traffic. Thanks for the replies~