Topic: [POOL][Scrypt][Scrypt-N][X11] Profit switching pool - wafflepool.com - page 261. (Read 465769 times)

But when are the payouts usually going out? I just started mining on Waffle.

No problemo 

Were resyncing a few things currently.  Payouts for today will be delayed slightly 3-4hrs from now), sorry for the delay.

+1

When do payouts occur? I am in an Eastern Time Zone

count me in!

Before disconnecting - tell us in 30 minutes. if disconnected, disconnect access to us that we were not working in idle. Thank you for your work!

Shit happens  

Happy to see that poolwaffle always deals with problems in a fast way and makes sure everything will work better after a fix!

on BTC u are right,
on alt-coins not.

i got a lot of problems with wafflepool and middlepool as failover.

every Blockfound from an other Pool with other coin make cgminer confuses

This is the best advice.  While I'd love to have 100% uptime, while wafflepool is growing like this, we are going to have some outages as things break.  I'll try my best to make sure what happened earlier doesn't happen again (server accepting connections/shares and not reporting them), and instead goes straight offline.  But its on you guys to make sure you have a (or two) backup pools enabled.  Just a good habit when you're using any pool

Yep, thats one disadvantage of openness   If/when things happen, I'm here to hear about it (and you guys can see it happen more), whereas with some other pools that don't share as much info publicly, it might happen often, just hidden from sight.


That data isn't really used for anything (other than listing for user sake).  It has been disabled for now (we'll see if it comes back - not sure yet).  The issue with it has to do with the way we're submitting blocks, and the way we're saving shares.

Shares get sent from endpoints to a collecting server (so they're all processed the same regardless of which endpoint you're on), and at the time that they're sent to the share processor, we don't know if they're a valid block yet.  We have a guess, but there are plenty of reasons the daemon might reject it (new block just came in, invalid data for some reason, etc), so we can't mark it as accepted on our side when we send to the share processor.  We then submit it to the blockchain, and know if its a valid block.  This was unfortunately one of the many things slowing down our original implementations.  Once we know it was accepted, we could go back and mark the previous share as an accepted block, but the share is buried in a pile of shares that are being actively processed.  It may have been processed already, it might still be waiting, so changing the flag to "accepted" requires that the queue be paused from additions, processed completely, update the share record, and reactivated.  The waiting for share queue to flush completely was causing the block finding server instance to lag a bit for upwards of 2-3 seconds.

Based on the fact that we really don't use it anywhere (we don't play extra to the block finder), during the mass changes a couple days ago I just went "too much to deal with right now", and removed the updating for now.

I can see what it will take to add it back in, if people think it is important.  One simple answer is to just queue up one additional share (the same share once with a "not a block" flag, once with "valid block" flag).  It makes accepted shares differ from what cgminer/etc shows, but not by a terrible amount.  Also could potentially make that fake share worth more as a block finders fee (fake share difficulty worth 1% of current shift total shares or something).


All that said.  I think the next step for the pool (other than making sure things don't break again), is adding per-worker stats to the miner stats page.  As well as going through some of the new coins to be added.

I looked at a few new coins to add, but either they've done some odd things that would require major upgrades to our server to support (odd transaction messages, etc), or the coin daemon isn't stable (cosmoscoin), or they're not supported on an exchange with the APIs we need.  I've reached out to both coins-e, and coinmarket.io about adding some APIs we would need to be able to auto-trade on their exchanges.  No progress there yet though.

Everyone should have a fail-over in-case of emergency set anyway... basic mining principle.

Issues like lost shares do happen (probably a fair bit on places like middlecoin that just pause stats for ages...), but poolwaffle always seem to be on top of them for wafflepool.

But, PoolWaffle, why are you Blocks Found stats blank?  Data not lost I hope?

good idea off, but better - Announce disable certain groups of servers or entire pool. will switch to standby. Thank you.

on faq it says on 0.01 we will be paid
but now ive 0.06159006 btc conformed and not paid...
is it time specific?

Sorry guys, that whole post was written before the emergency maintenance, and I wasn't able to hit send since the emergency came up.  I hate to have to pull the entire pool down, but our database started exhibiting some issues, and I needed to pull everything offline to make sure we didn't lose any data (and had consistent state to verify).

Everything is back up and running at this time, and we shouldn't need to do this again.  Sorry again.

As you're all aware, we had some issues with the eu endpoint dropping shares for a few hours yesterday (they were fixed as soon as it was noticed, and won't happen again), however the shares during that 3-4hr period are essentially permanently lost.  During the issues we were having a few days ago, we changed the way our shares are routed from endpoint boxes to our collector box (share processor).  Essentially this connection dropped yesterday, and the reconnect logic was never able to reconnect.  Two things have been fixed to stop this from happening again.  First, the reconnect logic is fixed (so if the connection does drop, it will be brought back up immediately), second, if the reconnect logic _does_ fail twice in a row, the server will shut down to allow miners to fall back to their backup pools.

As for dealing with the lost shares, we essentially have no way of knowing which shares (or how many) were lost during that time.  The only thing I can offer as a token is setting the pool fee lower for a time to make up for the lost shares..  I'll figure out how best to do this (it will cause some code issues, as blocks that have been processed as immature during 1% fee will be processed again at 0% fee, and will move 1% more unconfirmed --> confirmed than actually exist in unconfirmed, will need to be scheduled for a specific time in the future).



This doesn't even make sense.  First, why in the world would I fake hashrate?  What does that gain?  Second, if I were to fake hashrate, we wouldn't be earning any more coins than usual (and the block find records are listed right on the site), the profitability would absolutely plummet.


Contact is listed at the bottom of the site (footer), has been there since day 1 with an email address and a link to here.


EU is fine now, and in the next few days we will be looking into adding an extra endpoint in Singapore for you SEA guys

pool online again... useast working fine for now 

didn't try eu server yet...

eu server issue fixed?

let's listen excuses, whys and whys from the dev...

We're down for a bit of maintenance (sorry, shouldn't be much longer). Mining has been temporarily suspended.


site is down

Cgminer automatically switched to my back-up middlecoin. Apparently, everything is down at wafflepool at the moment. Also the front-end isn't reachable due to maintenance. Correct?