This isnt 100MB/s either. It wouldnt saturate bandwidth of 100MB/s either less you downloaded 100 files at the sametime of 2MB/s each
I’m not going through the 27 million files that the node is currently transferring just to prove my point. Travelling at 100km/h for 5 minutes is the same speed as travelling at 100km/h for 10 hours.
You cannot compared distance traveled with speed of internet… If you do a speed test of a file that is 2MB you cant benchmark your internet speed because it doesnt have time to ramp up.
1 Like
Ok I’ll humor you. How long would a 2MB file take to download if you downloaded it at 100MB/s?
It takes 0secs to download
No it doesn’t. It takes 0.2s to transfer. 0.2 != 0
Edit: why didn’t anyone catch my math error? 
0.02s
It actually takes 0 and it doesnt ramp up to anything
Sorry bud but your wrong. It barely goes over 17Mbps
Here is a 7MB file
No where near 100MB/s We can sit here all day and you will still be wrong. Internet doesnt just start at max speed… Just like you cannot download a 2MB file at 1000MB/s
Let’s clear something up. A file isn’t delayed if the bandwidth is there. Just because your sampling rate is too slow to see the ramp up/down, doesn’t mean that a 2M file is waiting for a second to be transferred at 2MB/s in order for it to take 1 second and register on a graph with a sampling rate of 1 second (ie how much bandwidth is used in 0s, 1s, 2s, 3s, and so on).
2 Likes
No you cannot download 2MB at 100MB/s you cant download at 1000MB/s
etc It doesnt just keep scaling up just because your internet is faster.
Guys your are both right. You cannot measure a speed for tiny files (with today’s channels not a 56kbit modem, right?), that’s it.
You need to test a transfer of a much bigger file, like 1GB at least to measure a speed.
Do we need mods to break out a separate thread titled “Graph Sampling Rates vs. Basic Math”? 
4 Likes
No I just wanted to make a point that a single file 2MB file cannot saturate 100MB/s internet speed.
right… but the speed depends on time and the sampling rate of your monitoring tool…
and we here again, do we need to split this (not equal or meaningful) comparison?
Sure if someone actually has the right tools to compare.
1 Like
Today my TrueNAS server is working hard xD
One of my nodes that usually averages 10GB/day egress just shot up 10x yesterday. I’m not complaining, bring on the egress!
Here is the benchmark with the right tools:
------------------------------------------------------------
Client connecting to 192.168.0.201, TCP port 5001
TCP window size: 16.0 KByte (default)
------------------------------------------------------------
[ 1] local 192.168.0.11 port 54162 connected with 192.168.0.201 port 5001 (icwnd/mss/irtt=14/1448/1619)
[ ID] Interval Transfer Bandwidth
[ 1] 0.0000-0.1137 sec 2.00 MBytes 148 Mbits/sec
1 Like
what is it?
please, elaborate