Author

Topic: [ANN] [QCN] QuazarCoin | Full privacy&data protection | Egalitarian PoW - page 104. (Read 186566 times)

newbie
Activity: 26
Merit: 0

Somethings not right with your pool. When I type in my address to check i get
Code:
[object Object]/sec
. And the maturity in the block list is missing (old codebase?)
full member
Activity: 224
Merit: 100
what's special/unique about this coin?
member
Activity: 85
Merit: 10
QuazarCoin will be a sensation of this summer!
hero member
Activity: 854
Merit: 539
Hello OP i have do the french transalation and send you a message can i have the bounty?

thanks
member
Activity: 90
Merit: 10
Myth
Another drk, compared drk technical advantage?
sr. member
Activity: 280
Merit: 250
sr. member
Activity: 364
Merit: 250
member
Activity: 90
Merit: 10
Myth
A huge difference on polo
full member
Activity: 168
Merit: 100
The mysterious
Greater innovation, will be a miracle birth?
member
Activity: 88
Merit: 10
luck
Wow, has been in progress, we need a bigger exchanges like bter.com. Allow more people to participate
full member
Activity: 224
Merit: 100
goog
Perfect emission curve, with top technology, go to the moon
newbie
Activity: 26
Merit: 0
I've enabled long-polling and all my miners instantly went booo, couldn't submit not even a share Sad And then switched back off again.

But thanks for sharing the script, I'll see where I can fit it in.

This pool software is just not behaving very well under the load..

The script is for miners to use - its a workaround for the pool bug (the pool bug hits even if long-polling is turned off), which means you can just ignore minerd and leave it to do its thing without worrying that its going "booooo" all the time instead of "yay!!!"

Otherwise, the only stable mining is using the daemon, or simpleminer (from the MRO source), and the latest AES-NI optimised version of cpuminer-multi performs better than both of them on cpu's which support AES-NI.


Yeah, I understand what's the script. However not sure how that bug works.

To what I experienced with my miners after enabling long-polling - I couldn't submit a single share, it went boooing straight from the start. Where as without long-polling the miner sometimes booos and then it's able to recover and carry on.

My next approach to the problem is to tune the share targeting time to as low as possible - 5-10s or even less. Shouldn't it decrease orphans as well?
member
Activity: 81
Merit: 10
<3 big picture
I've enabled long-polling and all my miners instantly went booo, couldn't submit not even a share Sad And then switched back off again.

But thanks for sharing the script, I'll see where I can fit it in.

This pool software is just not behaving very well under the load..

The script is for miners to use - its a workaround for the pool bug (the pool bug hits even if long-polling is turned off), which means you can just ignore minerd and leave it to do its thing without worrying that its going "booooo" all the time instead of "yay!!!"

Otherwise, the only stable mining is using the daemon, or simpleminer (from the MRO source), and the latest AES-NI optimised version of cpuminer-multi performs better than both of them on cpu's which support AES-NI.
newbie
Activity: 26
Merit: 0
I've enabled long-polling and all my miners instantly went booo, couldn't submit not even a share Sad And then switched back off again.

But thanks for sharing the script, I'll see where I can fit it in.

This pool software is just not behaving very well under the load..
member
Activity: 81
Merit: 10
<3 big picture

I've heard from others that long-polling causes more trouble than benefit. But yeah, ridiculous number of orphans. I'll give it a go and watch.

Actually it works pretty well - the problem we've got is the node-cryptonote-pool software has a significant bug with difficulty increases.

I'm running an internal pool which I was using to solo-mine yesterday with some amazon ec2 instances so I was having a play with various settings.

I posted this shell script in the MRO thread earlier - but its actually one I'm using to restart minerd when the difficulty bug strikes and I get "boooo" all the time..

Code:
#!/bin/bash

coin=QCN
pool=qcn.cryptity.com:7779
wallet=1234blahblah4321
minerpath="/home/ubuntu/minerd"
minerdcmd="minerd -a cryptonight -o $pool -u $wallet -p x -q"
maxboos=2
looptimer=15

while true; do
        running=`ps -C minerd | grep minerd | wc -l`
        maxnum=`cat /proc/cpuinfo | grep processor | wc -l`

        # check minerd is running
        if [ $running -lt 1 ]; then
                echo `date +%H:%M:%S` - $0 - No minerd found: starting miner on $coin @ $pool
                $minerpath/$minerdcmd  2>> $minerpath/minerd.log &
        # sanity check to make sure no more than one is running
        elif [ $running -gt 1 ]; then
                echo `date +%H:%M:%S` -  Too many minerd found: killing and restarting
                killall minerd && sleep 5
                $minerpath/$minerdcmd  2>> $minerpath/minerd.log &
        fi

        # check for boooo in the minerd output
        boos=`grep 'boooo' minerd.log | wc -l`
        #if there's fewer than allowed, just say so
        if [ $boos -lt $maxboos ]; then
                echo `date +%H:%M:%S` - Found $boos / $maxboos boos
                sleep $looptimer
        else
        #otherwise, there's too many, so shoot the miner
                rm minerd.log
                echo `date +%H:%M:%S` - FOUND TOO MANY BOOS - $boos / $maxboos
                #restart minerd
                killall minerd && sleep 5
                echo `date +%H:%M:%S` - Restarting miner on $coin @ $pool
                $minerpath/$minerdcmd  2>> $minerpath/minerd.log &
        fi
done;


It just checks to make sure minerd hasn't crashed, and otherwise watches for "booooo" in the output, at which point it restarts the miner.

Long-polling isn't new - and its well used by other currencies and cpuminer - the problem is the pool software is immature.. but long-polling saves a lot of the alternate-branch stuff.

QCN: 1Rz2t9dKzsKC6weRuxt7wESDWbx8BdHHJ7kFrYsVUc6E4kpMdzV5Yui8PGzwrF9kGu41RYky2wFN2CV YTFjBKuHJ4eQv4EL
MRO: 4BJgMQLr4d3XwqxzZsQrmvXHF2RXXGWuJSDqhtyQUebA2zz4G4Xa32S4rp8UTMb9kh2cVPCwoGZfUMT KvtFpBFN6AaU6kYR
newbie
Activity: 26
Merit: 0
I started a new mining pool: http://qcn.cryptity.com
Please give it a try.

You should enable long-polling, it will reduce the number of orphans a lot.

In your config.json, find this bit and change to true:

Code:
    "longPolling": {
        "enabled": false,
        "timeout": 8500
    },



I've heard from others that long-polling causes more trouble than benefit. But yeah, ridiculous number of orphans. I'll give it a go and watch.
legendary
Activity: 1792
Merit: 1010
also http://qcn.cryptity.com/

is stuck on

Last Share Submitted: 3 minutes ago

and 0 QNC payout last 30 minutes or so... nooot good
legendary
Activity: 1792
Merit: 1010
ooo waa....

transfer             transfer [ ... ] [payment_id] - Transfer ,... to ,... , respectively. is the number of transactions yours is indistinguishable from (from 0 to maximum available)

is the number of transactions yours is indistinguishable from (from 0 to maximum available)

juuuust excellent 
member
Activity: 81
Merit: 10
<3 big picture
I started a new mining pool: http://qcn.cryptity.com
Please give it a try.

You should enable long-polling, it will reduce the number of orphans a lot.

In your config.json, find this bit and change to true:

Code:
    "longPolling": {
        "enabled": false,
        "timeout": 8500
    },

Jump to: