0
Under review

Network Usage - Dash Pool

Jeffefe 6 years ago updated by Oliver 6 years ago 13

My 4x miners have been using 100 mb/s connecting to your Dash pool for the past week.


From what I've read this would happen if the stratum server was handing out too many shares.


I have since changed pools because I can't saturate the network link.

Blue is incoming traffic, Red is outgoing.


To be honest, what it looks like from that graph is that I'm requesting a bunch of shares, hearing no reply, and then receiving them all at once.

Are you sure that this is only the miners <-> pool traffic? No other connections / problems?

Im actually writing this tool https://feedback.coinfoundry.org/forums/1-general/topics/556-important-hash-power-splitting/ and at the moment I´m doing all my tests with DASH on this pool. Your traffic peaks/timings/values have nothing to do with what I´m monitoring...


I wouldn't say it's 100% pool -> miner traffic. I got this graph from my network engineer while they were troubleshooting our network speed and blamed it on me...


The vast majority is in fact going to poolmining.org though.

Actually no idea and my English isn´t the best...

Maybe CPU Mining with ASIC Difficulty... ? - Nonsens... ;-)

If the tool is ready we will answer such questions within minutes...

Give Oliver your Dash address... he has some more possibilities than you and me to watch what and when you delivered.

How do I get in touch with Oliver?

Under review

@Jeffefe: When you create a new post there's tab switch at the top where you can chose "Private".

@Jeffefe: By the way. You traffic analysis is not correct. The screenshot below shows performance metrics of the pool server responsible for the dash pool and a couple others for the past 7 days. Not even all miners combined push the amount of traffic you've mentioned. Far from it. The stratum protocol is very efficient.



+1

Thanks for looking into this Oliver. I'll let you know what I ended up finding out from the engineers. They're capturing some packets tonight to try and figure out what might be causing the bandwidth problem.


This lends some credence to my suspicion that it wasn't the miners after all though.

Out of curiosity. Did you find the culprit?

@Oliver - I just found out that the problem ended up being a malfunctioning WiFi access point that gave the illusion of slow network performance - and yet I got the blame because during the time period the techs analyzed, the traffic to dash.poolmining.org was the ONLY network traffic, nobody looked at it in context though to see how little traffic it really was. Funny how that works.

I suspected something like that. The stratum protocol consumes so few bytes that you'd need to push hundreds of updates per second to all miners to reach those numbers.