Topic: [1050 TH] BitMinter.com [1% PPLNS,Pays TxFees +MergedMining,Stratum,GBT,vardiff] - page 57. (Read 837101 times)

Think it might be that ^ post that you read? I think that answers my question
Simple search did the job.

I read some place you can lose the work.   but I am not sure if it was here.   one place that does say you lose the work is www.wemineltc     but they are a ltc pool.  I do think I have read it in this thread but this is a very long thread.

That's exactly what I have been doing. But I'm just curious

do not delete workers like that as there  no need to do it.

make a new worker move to the new one then delete the old worker the next day.

Just wondering, what will happen if I get a share accepted with my worker, then delete the worker immediately after?
Will the amount for the share be lost?

yes ant miner s2's are price in usd and btc is going up.  this makes ant miner s2's a better deal today over yesterday by serious difference    .


 also do you have an s2 coupon?  if you do that is 400 usd off the price.

       I would get s2's asap.

pm me if you need an s2 coupon as I have 8 ..

 I am off loading all my s-1's to get an s2 or 2 .

Ok...here it goes...

I am currently running 1.2TH in antminer S1's.  For a variety of reasons (all legal) I have been presented with a situation where my electricity costs are minimal.  (less than 6 cents kWh) The catch is that I only have 24U in a data rack with which to put miners in.  The room is also air conditioned and sound is not a problem. So...the only limiting factor is CapEx (I'm no lottery winner) and space.   It's funny you should mention Spondoolies because it would seem to solve my problem nicely.  Except...I have found the PepperMining Habanero.  I have run the numbers and I can fit 2 boards in a single re-purposed 2U server case complete with psu's and plenty of cooling, for a little less than half the CapEx of the Spondoolies. So 1.4TH vs 1.3TH in roughly the same space with a significant cost difference, minus the intangibles. (which might make a huge difference and I'm just too new to see them - i.e. who can you trust, who will be supportive, who's going to be around in 6 months time)

My question is...without being burdened by power efficiency, cooling or noise pollution, am I crazy to consider building up the Habanero's (which haven't even be produced yet) in lieu of going with a tried and tested Spondoolies (which is so highly priced that the ROI is scary, and delivery data is kinda iffy)?  Or I could simply stick with the tried and tested Antminer S2 which is shipping in 48hrs and jam as many as I can in the cabinet and call it a day. (a bird in the hand)

Or am I missing something so obvious that you're all screaming at your monitors right now?

The pools' collected wisdom is very welcome

Actually it's at 99.25% right now

Really, it's just bad luck. Bitminter has a link to a Wikipedia article explaining CDF in their Statistics -> Luck link that explains the concept behind the luck calculation and how it works.

whole pool*
Posted from Bitcointa.lk - #618i8obDf7LOIsrx

it is just me or is the whole clan stuck at CDF 99.24% AND 2days and 6 hours?  what's going on?
Posted from Bitcointa.lk - #wzlrVTmurjbRQHjb

 feel free to fire away.

I am been playing the btc game for about 2 years. 

First off.  BTC/any coin at all can crash and burn leaving you with nothing. So do not spend more then any money you can afford to turn into $0.00

Second off.  No piece of gear right now is 2x better then any other gear.

 What does this mean the best watt to hash  number is about .7 watts for 1gh hash
a spoondollies sp10 in low power mode  uses 700 watts for 1th.

most gear now is 1 watt to 2 watts per hash. 

Third fast delivery is very important.

Never hurts to ask.  This group has been very good at helping and sharing knowledge.   

Ok...that all makes sense. Thanks.

By the way, I have a general question about hardware and frankly after reading a lot of the posts over the past couple of weeks in that section, I'm flummoxed/discombobulated.  I know this is off topic but the guys(gals?) in this thread seem to be a bit more mature and considered in their responses...so shall I ask it here or maybe pm some of the guys?

I just don't want to be an ass...as well as a noob...

While you're right on luck affecting payouts, the fact remains that the *expected* payout for a share doesn't change from the moment it's submitted, and it's independent of any previous/future mining activity.

Hot/cold streaks are not possible to predict.  Basing a decision to join/leave based on past results is no different than gambling.  Since PPLNS pays shares based on future performance, it is not possible to hop in/out of luck with ANY reliability, which is why it's unhoppable.

all correct .  there is luck involved in the ramp up and ramp down.    As you ramp up  all hit blocks pay a little more    then the stable number as you maxed out.  when you leave and ramp down all blocks hit pay less.

so leaving in a cold streak is fine as long as the streak stays cold right until you have zero hash.  leaving in a cold streak is terrible if the pool gets really hot and you have very little hash left.

I think I will boost my hash now as a 49 hour streak is long time.

PPLNS has a charge-up and wind-down period.  When you first start, your reward is lower than what you'd expect per shift, until you have shares in all the shifts which are receiving payments.

When you stop mining, you continue getting paid until all the shifts you had shares in have left the window that gets paid.


However, the two periods are symmetrical.  You do not gain or lose anything by continuing to mine or leaving the pool.  At any given point, the *expected value* of any share submitted is completely independent of how long you have been mining previously, or how long you continue to mine on a PPLNS pool.

Sooo...I'm a noob...but...doesn't the PPLNS mean that when folks bail on the pool...it's going to take them a while to get back up to their previous share if they decide to come back when this block is solved and we get back to "normal"?

i pointed miner also at other pool.

long time only work and no payments.

Two days no pay, GigaBit don't stay.

Doc,
There are some issues with blacklotus
I have rebooted my miners - 20 tplinks

Half of them connected just fine and working,
the other half are connecting to backup

cgminer + configs is equal to all of them


Is there some sort of connection limit per IP? I think yes. So it turns out that it is not enough for me...
Thanks

The issue:

 .blacklotus.net (Or nginx proxy or whatever is there ) delays connection up to 120 secs

From other hand in cgminer code curl timeout value is set to 60 secs

So that was the core reason for issues

json_t *json_rpc_call(CURL *curl, const char *url,
            const char *userpass, const char *rpc_req,
            bool probe, bool longpoll, int *rolltime,
            struct pool *pool, bool share)
{
   
long timeout = longpoll ? (60 * 60) : 60;

And this on sockets

      FD_ZERO(&rd);
      FD_SET(pool->sock, &rd);
      timeout.tv_sec = 90;
      timeout.tv_usec = 0;

      /* The protocol specifies that notify messages should be sent
       * every minute so if we fail to receive any for 90 seconds we
       * assume the connection has been dropped and treat this pool
       * as dead */
      if (!sock_full(pool) && (sel_ret = select(pool->sock + 1, &rd, NULL, NULL, &timeout)) < 1) {

I am damn sure that  .blacklotus.net (Or nginx proxy or whatever is there ) is causing this because i have many simultaneous connections originating from same IP


Please do one of two things

1. Ask blacklotus.net/Nginx to delay connections not more than 50 secs
2. Consider providing some kind of secret URL mining or VPN or whatever for your trusted users - to bypass blacklotus/Nginx

However first option is preferable