Topic: KanoPool kano.is lowest 0.9% fee 🐈 since 2014 - Worldwide - 2432 blocks - page 2019. (Read 5352086 times)

That being said, what is your ping loss on a continuous ping to stratum.kano.is?

Not sure if I cut this post down to size correctly but I'm trying to help out this user. I will first say, you are on the right track but I don't think you will accomplish anything much but lose sleep. I am running Verizon FIOS here (fiber optic to the house) and still drop many a packet. When I questioned my provider, I was told that this was normal behavior as they customarily restart network flow to ensure that old connections aren't draining system resources...yada, yada. All I am stressing to you is this...don't chase a ghost! If you wish to proceed, by all means & methods, I wish you luck. I hope your hair can stand the pulling.

Good Luck & Happy Mining

     

I second that

closely followed by 2 blocks of 5%...... i dream 

100% block

Arris gateways were known to have dropping issues when used as a home gateway. I recall something about changing the DHCP scope away from 0.2 -0.9, as those addresses seemed to be most affected. I think there was also an issue with high active/open connections causing them to overflow and reset the network stack. (this was many years ago - I would have assumed it would have been addressed by now)

The Arris gateways aren't really designed to be run as complete network gateways. They're best used as a bridge to a dedicated router if stability is something you are looking for. I'm still running an old 54G as a wireless bridge in front of the miners and use an RT-N60 for my main router.

On the graph:
Big one is S5+ single unit
S4 single unit
SP20 multiple units in operation single unit shown
S5 same as SP20

I am thinking network because I am not pulling 50 amps continuously and I am sized well under 70%. #2 copper runs from the main to the house to a 125 amp sub panel. It is ran inside liquidtight. Good stuff. I have checked all connections and measured solid line voltage, always slightly higher by a small percentage. No fluxuations my fluke can pick up, and it isn't a scope so there is a ton you cant see with a meter.




I am leaning towards net, I loaded a couple of things like wireshark but I haven't seen anything obvious. I will need to be more familiar with the software and how to configure it, then it will be most useful.
I want to record packet flow at my router right?

S3+ could be singled out in that chart as well and looks the same way, most of the big spikes line up. but I think it shows there is a definite issue being magnified on larger miners.
Harmonics?
It is the above listed minor small nano farm on the generic wire shelving you see in many pics. I use several 2880 power supplies with b/o boards from Finsky and Jabber. I supplement using DSP2000BB PSUs with top shelf b/o boards from sidehack and novac of Gekkoscience. Solid work, and they can take some stress if you have a power supply down. You can push either one of these hard if you need to do so. I do not recommend exceeding any specifications.

I am interrupting the thread with hardware troubleshooting.

Thank you fellas for the replies. I will take a stroll through my connections again and swap some cables, think simple for a bit and proceed.

Much Appreciated and thanks for the solid pool, I'll get my connection solid as well.

It DOES.

Your pool better pay better than Eligius ;-) I would say ;-) /cheers ...

Edit: it will not ;-) BUT your Reply/Views ratio is impressive in this thread ;-) so I will stay at Kano.is ;-)

Well all I could suggest is to run some sort of network monitoring to see if your network is dropping out.
I'm not a windows fan so I've no idea what to use for that.

I (or course) run a linux box (using pppoe) as my network connection at home so I see when my network drops out and reconnects - and that matches the times when my miners drop out and reconnect. I force this once a day, but it often happens at other times also.
In my case it's just the way it is with my ADSL connection (last night from 10pm to 4am it happened 6 times ... only one of those was my daily reconnect by design ... that was the worst number in a long time)

There could of course be other reasons, but e.g. if it was the miner dropping the connection you'd need to monitor that somehow in the miner.

Not all the spikes seem to apply at the same time. But some does, so it doesn't seem like a single unit, but not always the same either. I have that with my S5, maybe you have a combination thereof?

Maybe if you tell us what miners they are it could help shedding some light.

The big one is you blatantly losing connection for a few minutes however.

Kano, Thanks for explaining. I geek out thinking about how much work goes in running a pool, much less coding, debugging, and hitting back with great NFO.

I've had a problem, or think I do, with my internet dropping out quickly. It could be something else, maybe even incoming power.
The difficulty with troubleshooting with my meter comes from a limited high/low or in general record time.

Also, I have seen my cable / wifi router LEDs drop out. I'm not sure, it was only once recently I saw this. It is an Arris, one you rent for 8 bucks a month. It actually seems like decent hardware, but the firmware sucks and is limited.

Anyway, I noticed this behaviour (I think this shows what I mean) on the graph:



Seems like I lose the connection. Sometimes I refresh the workers speed and can tell some? of the workers are showing hashrate similar to what they do when connected for a short period of time. The clock shows the miner has been up, so it would look more connection related, maybe even rollovers because I do at times see work performed on my backup.

 I am sure there are tools or methods maybe some simple win7 commands to get me on the ball with troubleshooting this, or first properly identifying (helping me understand what the pool thinks is happening.

I haven't ruled anything out, but for incoming power to be an issue I would expect to see the up time displayed on the miner to be reset.

PS the hard work you do and the members in this thread throwing out knowledge is fantastic.

I had injections in my back this morning. Twas no fun really, meds are going pretty well, Just before and during the procedure I try not to think about what is going on I concentrate on the gorgeous ladies. My Doc loves being surrounded by beautiful women. Such a tough life. He is a good guy though. He knows exactly what to try next, or not try and ride it out for a bit. Sometimes less is absolutely best.

Payout 373960 sent
193d5d963b002d07dc971b03731e8a9533e9427b0a6518b46734ed19a8607dc9
and confirmed

ckpool normally sends out new work (a workinfo) every 30s
It also sends out extra work, every network block change, but the next work change after that will be 30s after the previous work change, so effectively that is an extra workinfo - so each network block in a shift represents ~30s less for that shift

A normal shift is 100 workinfos.
A shift finishes early (before 100) if either:
1) We find a block
2) I restart ckpool

So the maximum expected shift length is 50minutes, but all the block changes reduce that.

The only effect the length of a shift has on the payout, is due to the length of the first shift in a payout range.
When we find a block, ckdb counts back in time, all the shares in the shifts until it equals or exceeds 5Nd
Since it can't divide up a shift, it must include each full shift up to that point.
Thus the shift at the start of the payout range, will, on average, include an extra half of that shift of shares.

Thus the payout covers 5Nd plus a bit, so that means shares, on average, will get rewarded a small bit more than 5 times and thus shares will be rewarded a small bit less than 1/5th of their expected value, each time they are rewarded.

My miners were off for a while (network switch died) and I brought them back online Sept 4th. I knew that it would take a while to back up to full payout level, but I just looked through the shifts and figured out I still have about another 2.8 days before the zeros are outside the 5N range. 

Kano, quick question I'm sure has been answered but I either missed it or forgot:
How are the length of the shifts determined, I understand the shorter ones when we find a block, but the others are around 45 minutes, but all different?

I will add that the overall effect is not to punish either.

If blocks were always 100%:
If you stopped mining for 1 day, then for the N Range after that (~9 days) you would get 8/9ths of your reward.
If the N range was one day, then for those 9 days, you'd have one day of no reward and 8 days of normal reward.
Works out the same when luck is 100%.

When luck is above 100%, your total extra is of course also affected with the ~9 day N Range, you are getting more, but not 100% more, instead 8/9ths more each time since you weren't mining the whole time.

The simple answer is, the average expected income of mining half the time is half the reward.
Replace half, with whatever % you mine.
If it's simply just an outage, then look at the (currently) 9 days after the outage and expect to get, on average, that % of the 9 days you were mining.

What actually happens, is of course controlled by ... luck ... since we don't get 100% blocks every time ... and have averaged better than 100% over the past year

I kinda expected that explaination....damn!

Thanks for the time spent.

'nuff Said
     

For both of you above.

During the range of each reward, you have a gap of no shares.
Until that gap is no longer in the payout range, your block reward will be affected by it.

It's not a penalty by design, it's simply that the good pool luck means it is affecting more payouts, since there are more payouts in the 5Nd range from the start of your gap of no mining.

So what is happening is you are getting extra rewards, but of course they are effected by the gap in the N Diff/N Range of each reward.
i.e. you are being paid more often than expected, but getting a bit less while the gap is in the reward range.

Lady luck is a harsh mistress, she's paying you more often but giving you a bit less extra each time

Hello Guys !! For one day I had my miners turned off and my hashrate drop from 3,9 TH to 3,38. After one day I turned on my hardware but since shutting down "My N Avg" stat is not rising - 3,38 TH - it's still the same even though my miners are working with full hashrate. What should I do? Everytime I'm earning almost 0.01 btc less for block.

@ kano,

Will you please explain why my reward N avg is not going back up to the 1+TH range? I know I've had some electrical issues earlier this month causing me to drop my hashrate, but, as of late, my equipment has been performing as expected. I'm not understanding why the N avg hasn't gone back up.

Thanks,

'nuff Said

     

They accumulate.
Will be paid some time ... in the future when I add payout confirmation to ckdb and thus have an unpaid balance.

Indeed if you are concerned about getting dust soon, don't mine here.

You can already see your total reward at the bottom of the reward page.
You can work out your unpaid balance by subtracting your wallet balance on your payout address - assuming you only use the address for pool payouts.