[1050 TH] BitMinter.com [1% PPLNS,Pays TxFees +MergedMining,Stratum,GBT,vardiff] - page 247.

DrHaribo

legendary

Activity: 2730

Merit: 1034

Needs more jiggawatts

Quote from: Bitsilver.us on February 25, 2013, 09:15:37 AM

My apologies if you have already covered the subject, but ... what are your commits on the use of mining_proxy with the new ASIC boxes coming online?

And does anyone know if bfgminer supports the variable difficulty being talked about here?

My recommendation is to not use proxies unless you actually need them. Bfgminer has Stratum and vardiff support built in, so you'd be better off connecting it directly to the pool. But if you have multiple mining computers on a local network, you may save a little bandwidth by using a Stratum proxy instead of having each connect to the pool directly.

Epoch

legendary

Activity: 922

Merit: 1003

Quote from: Bitsilver.us on February 25, 2013, 11:48:12 AM

Quote from: Epoch on February 25, 2013, 10:20:06 AM

Quote from: Bitsilver.us on February 25, 2013, 09:15:37 AM

And does anyone know if bfgminer supports the variable difficulty being talked about here?

Yes, for the past several versions. Your worker hashrate needs to be high enough, though. Roughly 20 shares/minute is the threshold I believe. I have a 1.6Ghps worker (~22 shares/minute) which never sees a difficulty higher than 1. A 12Ghps worker would see a difficulty of 8.

We have one bfgminer setup with 24Ghps total, so I'm assuming we should see the difficulty rising. What this quickest way for me to check on that, and is the difficulty level something I can set myself?

Thanks for your feedback...

To check, just connect your miner to the pool (-o http://mint.bitminter.com:8332 -u -p ) and wait; whether you use stratum or not doesn't matter. It will start at difficulty 1 but in a short while (a minute or two) you will see the difficulty increase as the pool gets a better idea of your miner's speed. BFGMiner displays the target difficulty in real time as shares are submitted, so once you see that number change from '1', you know vardiff is working for you.

Something like this (the 'Diff 32/8' means that the target difficulty is 8 and the submitted share had a difficulty of 32):

Code:

[2013-02-24 21:19:11] Accepted 08e39d0a BFL69 pool 0 Diff 32/8

In your case with a speed of 24Ghps you may see a difficulty of 16. As for setting this yourself, no, currently that is not possible at BitMinter. It is automatically adjusted by the pool based on your hashrate.

Bitsilver.us

newbie

Activity: 14

Merit: 0

Quote from: Epoch on February 25, 2013, 10:20:06 AM

Quote from: Bitsilver.us on February 25, 2013, 09:15:37 AM

And does anyone know if bfgminer supports the variable difficulty being talked about here?

Yes, for the past several versions. Your worker hashrate needs to be high enough, though. Roughly 20 shares/minute is the threshold I believe. I have a 1.6Ghps worker (~22 shares/minute) which never sees a difficulty higher than 1. A 12Ghps worker would see a difficulty of 8.

We have one bfgminer setup with 24Ghps total, so I'm assuming we should see the difficulty rising. What this quickest way for me to check on that, and is the difficulty level something I can set myself?

Thanks for your feedback...

Epoch

legendary

Activity: 922

Merit: 1003

Quote from: Bitsilver.us on February 25, 2013, 09:15:37 AM

And does anyone know if bfgminer supports the variable difficulty being talked about here?

Yes, for the past several versions. Your worker hashrate needs to be high enough, though. Roughly 20 shares/minute is the threshold I believe. I have a 1.6Ghps worker (~22 shares/minute) which never sees a difficulty higher than 1. A 12Ghps worker would see a difficulty of 8.

Bitsilver.us

newbie

Activity: 14

Merit: 0

My apologies if you have already covered the subject, but ... what are your commits on the use of mining_proxy with the new ASIC boxes coming online?

And does anyone know if bfgminer supports the variable difficulty being talked about here?

teflone

hero member

Activity: 770

Merit: 500

You're fat, because you dont have any pics on FB

Thank you

your work is appreciated...

DrHaribo

legendary

Activity: 2730

Merit: 1034

Needs more jiggawatts

Back up again. Sorry for the downtime.

bitcoind went haywire with disk accesses, but a restart cleared it up.

DrHaribo

legendary

Activity: 2730

Merit: 1034

Needs more jiggawatts

Having an issue with the server. Looking into it now.

philipma1957

legendary

Activity: 4326

Merit: 8899

'The right to privacy matters'

email notice will be nice. I am waiting on BFL to produce my 2x Jalapenos and 1x little single.

We have slowed our coins the 19 20 21 and 22 have been lite but the pool has gone up a bit. maybe the 23rd will be better.

coins dropped under 30 usd

DrHaribo

legendary

Activity: 2730

Merit: 1034

Needs more jiggawatts

Quote from: loshia on February 22, 2013, 03:39:54 PM

Var diff can slow things down a bit. Do you have plans to implement preconfigurable diff per worker in near feature? For 64300-64500 MHash/s diff about 64 shall be just right in order to have about 20 submited shares per minute and not to "flood" the pool right?

Bitminter adjusts the diff quite quickly. Asicminer tested the pool with much more hashrate than one Avalon and the pool adjusted (it maxed out at 64 at the time, but I have set a higher max now). That said, there are plans to allow setting a minimum diff.

Quote from: loshia on February 22, 2013, 03:39:54 PM

PS: As long some instability issues are reported from avalon users a nice feature available via perks will be inactivity mail notification - per worker. What about it?

Yep, that's also on the TODO list.

loshia

legendary

Activity: 1610

Merit: 1000

Doc,
I am waiting for my avalons to land. And i love your pool and i will mine here. However according to this thread

https://bitcointalksearch.org/topic/my-impression-of-avalon-machine-cooling-info-updated-145885

-----------------------------------
4 please DO preconfigure the diff of ALL the pools (including the backup ones), var diff may slow down the machine at first until the diff is adjusted.
-----------------------------------

Var diff can slow things down a bit. Do you have plans to implement preconfigurable diff per worker in near feature? For 64300-64500 MHash/s diff about 64 shall be just right in order to have about 20 submited shares per minute and not to "flood" the pool right?

Best

PS: As long some instability issues are reported from avalon users a nice feature available via perks will be inactivity mail notification - per worker. What about it?

DrHaribo

legendary

Activity: 2730

Merit: 1034

Needs more jiggawatts

Quote from: loshia on February 22, 2013, 10:34:23 AM

Can i share this with Kon after i check it?

Of course. Anonymize out your IP address if you want. I didn't do that as I was sending it only to your email address and it could be helpful to you.

loshia

legendary

Activity: 1610

Merit: 1000

Quote from: DrHaribo on February 22, 2013, 10:20:30 AM

Quote from: loshia on February 22, 2013, 10:00:15 AM

1.when i mine with getwork with diff 1 given by the pool i submit every share
2. When i mine with stratum and diff 8 given by the pool i submit one of eight at average - there is variation of course
3. Is it possible in theory that those 8 shares skipped to be able to solve the block? Because if i was using getwork with diff 1 given by the pool they shall be ok and submitted to the pool right?

If your proof of work hash is above difficulty 1.6 million then it is also above difficulty 8, so there won't be any problems. Any proof of work that can create a valid block will be submitted by the client.

Edit: I emailed the data that was logged about your proof of work to the email address registered with your Bitminter account.

10X doc
Can i share this with Kon after i check it?

DrHaribo

legendary

Activity: 2730

Merit: 1034

Needs more jiggawatts

Quote from: loshia on February 22, 2013, 10:00:15 AM

1.when i mine with getwork with diff 1 given by the pool i submit every share
2. When i mine with stratum and diff 8 given by the pool i submit one of eight at average - there is variation of course
3. Is it possible in theory that those 8 shares skipped to be able to solve the block? Because if i was using getwork with diff 1 given by the pool they shall be ok and submitted to the pool right?

If your proof of work hash is above difficulty 1.6 million then it is also above difficulty 8, so there won't be any problems. Any proof of work that can create a valid block will be submitted by the client.

Edit: I emailed the data that was logged about your proof of work to the email address registered with your Bitminter account.

loshia

legendary

Activity: 1610

Merit: 1000

Quote from: DrHaribo on February 22, 2013, 09:30:28 AM

Quote from: loshia on February 22, 2013, 06:50:57 AM

Recently i found a block but cgminer did not report it. I have found blocks before with getwork and i was able to see Found block count > 0 matching the actual found blocks reported buy bitminter

Maybe this is a bug in cgminer. Or if you have multiple workers, did you check them all?

Quote from: loshia on February 22, 2013, 06:50:57 AM

1. Mining software knows that it shall shall submit work only if its matching difficulty (let say 4).
2. Mining software shall check if valid shares which are less that required difficulty solve the block

You mean above the required difficulty to solve the block. Above difficulty = below target.

Quote from: loshia on February 22, 2013, 06:50:57 AM

3. If yes they shall be submitted to the pool even though they are not matching pool difficulty

The difficulty given by the pool (1-4096 currently) will always be lower than the difficulty for creating a block (currently 3.6 million).

Even if the mining difficulty was not capped (currently at 4096) you would still need a worker with a hashrate that would produce 20 bitcoin blocks per minute on average, for the server to give you a mining difficulty that is higher than the difficulty to create a block. If you receive enough ASICs for that then I recommend splitting them up over multiple worker accounts, at least until the network difficulty goes up.

Quote from: loshia on February 22, 2013, 06:50:57 AM

What i am asking is are any potential issues from both miner and pool if a share solving the block can be rejected by the pool or not sent by software?

It could be a potential problem if you added a new worker that had 200 times the hashpower of the rest of the network. So if the current network hashrate is 30 TH/s and you add hardware doing 6000 TH/s on a single worker account, then we need to look into it. In that case, please let me know in advance. Wink

Thanks doc!

My explanation sucks.
What i am talking about is:

1.when i mine with getwork with diff 1 given by the pool i submit every share
2. When i mine with stratum and diff 8 given by the pool i submit one of eight at average - there is variation of course
3. Is it possible in theory that those 8 shares skipped to be able to solve the block? Because if i was using getwork with diff 1 given by the pool they shall be ok and submitted to the pool right?

Yes i do have two workers and checked them both - none of them was showing that block was fond and no i do not have asic yet:)

If you have log of share which solved block http://bitminter.com/block/btc/0000000000000201264143a1bca71fc300d02bf0a3e0d8cd85a407024abd7129
please PM it to me
10X

DrHaribo

legendary

Activity: 2730

Merit: 1034

Needs more jiggawatts

Quote from: loshia on February 22, 2013, 06:50:57 AM

Recently i found a block but cgminer did not report it. I have found blocks before with getwork and i was able to see Found block count > 0 matching the actual found blocks reported buy bitminter

Maybe this is a bug in cgminer. Or if you have multiple workers, did you check them all?

Quote from: loshia on February 22, 2013, 06:50:57 AM

1. Mining software knows that it shall shall submit work only if its matching difficulty (let say 4).
2. Mining software shall check if valid shares which are less that required difficulty solve the block

You mean above the required difficulty to solve the block. Above difficulty = below target.

Quote from: loshia on February 22, 2013, 06:50:57 AM

3. If yes they shall be submitted to the pool even though they are not matching pool difficulty

The difficulty given by the pool (1-4096 currently) will always be lower than the difficulty for creating a block (currently 3.6 million).

Even if the mining difficulty was not capped (currently at 4096) you would still need a worker with a hashrate that would produce 20 bitcoin blocks per minute on average, for the server to give you a mining difficulty that is higher than the difficulty to create a block. If you receive enough ASICs for that then I recommend splitting them up over multiple worker accounts, at least until the network difficulty goes up.

Quote from: loshia on February 22, 2013, 06:50:57 AM

What i am asking is are any potential issues from both miner and pool if a share solving the block can be rejected by the pool or not sent by software?

It could be a potential problem if you added a new worker that had 200 times the hashpower of the rest of the network. So if the current network hashrate is 30 TH/s and you add hardware doing 6000 TH/s on a single worker account, then we need to look into it. In that case, please let me know in advance. Wink

loshia

legendary

Activity: 1610

Merit: 1000

Doc,

Recently i found a block but cgminer did not report it. I have found blocks before with getwork and i was able to see Found block count > 0 matching the actual found blocks reported buy bitminter
I am curious how do you check valid shares when diff >1? What i mean is this:

According to my understanding it works like the following:

1. Mining software knows that it shall shall submit work only if its matching difficulty (let say 4).
2. Mining software shall check if valid shares which are less that required difficulty solve the block
3. If yes they shall be submitted to the pool even though they are not matching pool difficulty
4. Pool shall check first (or at least before diff match ) if the share submitted solves the block and later on if it matches difficulty right?

What i am asking is are any potential issues from both miner and pool if a share solving the block can be rejected by the pool or not sent by software?
I am just curious:)
10X

PS: just for the record i wanna check why cgminer does not report solved block. If you do keep log of share solving the blocks and you can paste here the winning share for block
http://bitminter.com/block/btc/0000000000000201264143a1bca71fc300d02bf0a3e0d8cd85a407024abd7129

will be great

10X

Digigami

sr. member

Activity: 462

Merit: 250

Working fine over here..

dooferorg

full member

Activity: 163

Merit: 100

Quote from: DrHaribo on February 19, 2013, 04:46:17 PM

Quote from: dooferorg on February 19, 2013, 04:34:06 PM

Hmm.. I had cgminer 2.10.2 and upgraded to 2.10.5 and still get a lot of rejects (about 9%), although not from 'below difficulty' as an error, just 'rejects'. Running for a day with just the regular getwork server via the nginx/PHP proxy I use and I get nearly zero rejects. I will try again with direct from miner to the stratum proxy and see if it's my nginx/PHP proxy getting in the way. Pity it if is since it's SO useful for pool failovers and configurations and monitoring all miners at once.

Yeah, 9% is ridiculous. Something is wrong there. I guess somewhere in your dual proxy setup things get slowed down a lot. Try connecting cgminer directly to the pool and see how that works.

Poolwide rejects for the current round is 0.16% as I write this. I'd expect some getwork users are above and most Stratum users are below that.

Yea, I had wondered if my setup was confusing things.. Seems like it is. Sad

Telling cgminer 2.10.5 to go to stratum+tcp://mint.bitminter.com:3333/ yields zero rejects for the 10 minutes I've been running it. Before I would see the rejects pop up now and then a few times by now. Well, at least I'll have my crap together by the time BFL ships.. lol.

DrHaribo

legendary

Activity: 2730

Merit: 1034

Needs more jiggawatts

Quote from: dooferorg on February 19, 2013, 04:34:06 PM

Hmm.. I had cgminer 2.10.2 and upgraded to 2.10.5 and still get a lot of rejects (about 9%), although not from 'below difficulty' as an error, just 'rejects'. Running for a day with just the regular getwork server via the nginx/PHP proxy I use and I get nearly zero rejects. I will try again with direct from miner to the stratum proxy and see if it's my nginx/PHP proxy getting in the way. Pity it if is since it's SO useful for pool failovers and configurations and monitoring all miners at once.

Yeah, 9% is ridiculous. Something is wrong there. I guess somewhere in your dual proxy setup things get slowed down a lot. Try connecting cgminer directly to the pool and see how that works.

Poolwide rejects for the current round is 0.16% as I write this. I'd expect some getwork users are above and most Stratum users are below that.

Topic: [1050 TH] BitMinter.com [1% PPLNS,Pays TxFees +MergedMining,Stratum,GBT,vardiff] - page 247. (Read 837122 times)