[ANN][The Original Multipool - Scrypt/SHA256/Scrypt-N/X11] multipool.us

[kslavik]

I could only find PPCoin worth looking into for SHA 256 mining.

[1714980663]

1714980663

[1714980663]

1714980663

[1714980663]

1714980663

[flound1129]

Well the multiport just stopped working again..  Luckily I was around to restart it.

I think it may just be a race condition in the script.  I added a 2 second sleep between killing off the old redirect and starting up the new one, hopefully that will take care of the issue.

[Swimmer63]

Miners are are working but I'm not seeing any results on the site since restart.

[Swimmer63]

OK - coming back.

[ranlo]

Well the multiport just stopped working again..  Luckily I was around to restart it.

I think it may just be a race condition in the script.  I added a 2 second sleep between killing off the old redirect and starting up the new one, hopefully that will take care of the issue.

Thanks, .

I was also wondering, is it possible to lower the difficulty to 8 or even 4 for the coins with very fast blocks? I am getting 15-20% stales at the current setting of 16 since the blocks move so fast.

[flound1129]

Well the multiport just stopped working again..  Luckily I was around to restart it.

I think it may just be a race condition in the script.  I added a 2 second sleep between killing off the old redirect and starting up the new one, hopefully that will take care of the issue.

Thanks, .

I was also wondering, is it possible to lower the difficulty to 8 or even 4 for the coins with very fast blocks? I am getting 15-20% stales at the current setting of 16 since the blocks move so fast.

The profitability calculation already takes stales into account (25% for WDC, 20% for DGC).

I'd have to look into the performance implications of lowering it below 16.  It would result in double or quadrupling the number of shares inserted into the DB.

[ranlo]

Well the multiport just stopped working again..  Luckily I was around to restart it.

I think it may just be a race condition in the script.  I added a 2 second sleep between killing off the old redirect and starting up the new one, hopefully that will take care of the issue.

Thanks, .

I was also wondering, is it possible to lower the difficulty to 8 or even 4 for the coins with very fast blocks? I am getting 15-20% stales at the current setting of 16 since the blocks move so fast.

The profitability calculation already takes stales into account (25% for WDC, 20% for DGC).

I'd have to look into the performance implications of lowering it below 16.  It would result in double or quadrupling the number of shares inserted into the DB.

Ah, got ya. So it's not just me with tons of stales then... that makes me feel a lot better!

If it's too much of an issue don't worry about changing it to a lower value. I just figured an increase of 10% or so is pretty big when you consider it over time. For ex:

If you get 10 coins normally, now you get 11
After a month you've gotten 30-31 extra coins, or 3+ days worth as a bonus

[flound1129]

Well the multiport just stopped working again..  Luckily I was around to restart it.

I think it may just be a race condition in the script.  I added a 2 second sleep between killing off the old redirect and starting up the new one, hopefully that will take care of the issue.

Thanks, .

I was also wondering, is it possible to lower the difficulty to 8 or even 4 for the coins with very fast blocks? I am getting 15-20% stales at the current setting of 16 since the blocks move so fast.

The profitability calculation already takes stales into account (25% for WDC, 20% for DGC).

I'd have to look into the performance implications of lowering it below 16.  It would result in double or quadrupling the number of shares inserted into the DB.

Ah, got ya. So it's not just me with tons of stales then... that makes me feel a lot better!

If it's too much of an issue don't worry about changing it to a lower value. I just figured an increase of 10% or so is pretty big when you consider it over time. For ex:

If you get 10 coins normally, now you get 11
After a month you've gotten 30-31 extra coins, or 3+ days worth as a bonus

The pool efficiency is showing as ~90-95% though for WDC and DGC (slightly higher for DGC, as expected) and we're not getting tons of orphans..  How many stales you get will really depend on your miner's hashrate. 

[flound1129]

Well I just got the Amazon bill for last month.  $152..  And the instance we have is practically at its limits.

So, I'm soliciting donations now for a new server.  Looking at 8x2.5GHz, 32GB memory, and SSD for storage.  I figure we need about $2k to get something that will last 3 years or so and will allow me to add as many different coins as you guys want.

I'm planning to create a 'donor status' on the site and put a special icon next to people who have donated.  In the future, there will be additional benefits for donors.

[Moray]

I'm confused with the % of donations... It shows 0 ( or empty ), every time I log in.
Can you, please, check if it's 0 ?
User ID: 1172

[ISAWHIM]

How the heck do we setup multiport for scrypt and sha-256 at the same time. (Terracoins are sha-256 and run like crap with sha-256 settings... not to mention, cgminer is setup for scrypt for the login...)

Just curious, because that would result in possibly valid scrypt-hash-shares being returned if you check them for scrypt-validity, but would never result in a valid sha-256 block ever being found.

Or am I missing something, like a special tool that connects and auto-switches video-card profiles/settings and miner-types from scrypt to sha-256?

Personally, I would drop terracoins, because they are sha-256, and will soon be invaded by someone with an ambitious THs army, to pump-n-dump BTC that is failing them in the future. Thus, degrading any holdings and earnings for us GPU and CPU scrypt-miners.

You have a theme, stick with it. "Scrypt-Mining" Get every scrypt-mined coin that exists, to be included in your pool. (Once you get that new server setup.)

FYI: You can buffer and compound multiple DB updates to the same table, if you use temp-tables to "hoard" the micro-updates. (Talking about shares that are slightly smaller and more realistic to less-loss.)

Another trick is to use PHP's built-in MySQLite as the unhindered multi-table. Every Nth or EndRound, you make one update to the real database, which could be a remote DB sent just the table-update info. Since it is just WRITE/APPEND and not REPLACE/UPDATE, for share counting.

[ranlo]

How the heck do we setup multiport for scrypt and sha-256 at the same time. (Terracoins are sha-256 and run like crap with sha-256 settings... not to mention, cgminer is setup for scrypt for the login...)

Just curious, because that would result in possibly valid scrypt-hash-shares being returned if you check them for scrypt-validity, but would never result in a valid sha-256 block ever being found.

Or am I missing something, like a special tool that connects and auto-switches video-card profiles/settings and miner-types from scrypt to sha-256?

Personally, I would drop terracoins, because they are sha-256, and will soon be invaded by someone with an ambitious THs army, to pump-n-dump BTC that is failing them in the future. Thus, degrading any holdings and earnings for us GPU and CPU scrypt-miners.

You have a theme, stick with it. "Scrypt-Mining" Get every scrypt-mined coin that exists, to be included in your pool. (Once you get that new server setup.)

FYI: You can buffer and compound multiple DB updates to the same table, if you use temp-tables to "hoard" the micro-updates. (Talking about shares that are slightly smaller and more realistic to less-loss.)

Another trick is to use PHP's built-in MySQLite as the unhindered multi-table. Every Nth or EndRound, you make one update to the real database, which could be a remote DB sent just the table-update info. Since it is just WRITE/APPEND and not REPLACE/UPDATE, for share counting.

I was wondering about the TRC as well, when I looked through what the different coins were. I was also going to recommend removing it, although I think that maybe it's on the pools in general but not on the auto rotator?

[flound1129]

How the heck do we setup multiport for scrypt and sha-256 at the same time. (Terracoins are sha-256 and run like crap with sha-256 settings... not to mention, cgminer is setup for scrypt for the login...)

This is impossible.  You can only configure a miner instance to mine SHA256 or scrypt coins.  It can't switch between them.

Just curious, because that would result in possibly valid scrypt-hash-shares being returned if you check them for scrypt-validity, but would never result in a valid sha-256 block ever being found.

Or am I missing something, like a special tool that connects and auto-switches video-card profiles/settings and miner-types from scrypt to sha-256?

Personally, I would drop terracoins, because they are sha-256, and will soon be invaded by someone with an ambitious THs army, to pump-n-dump BTC that is failing them in the future. Thus, degrading any holdings and earnings for us GPU and CPU scrypt-miners.

You have a theme, stick with it. "Scrypt-Mining" Get every scrypt-mined coin that exists, to be included in your pool. (Once you get that new server setup.)

FYI: You can buffer and compound multiple DB updates to the same table, if you use temp-tables to "hoard" the micro-updates. (Talking about shares that are slightly smaller and more realistic to less-loss.)

Another trick is to use PHP's built-in MySQLite as the unhindered multi-table. Every Nth or EndRound, you make one update to the real database, which could be a remote DB sent just the table-update info. Since it is just WRITE/APPEND and not REPLACE/UPDATE, for share counting.

I have considered going all-scrypt, but we have some very dedicated TRC miners on the pool (it was one of the first coins I offered) and I do want to eventually include BTC once we have enough SHA256 hashrate.

As far as your db suggestions, I agree that there could be much improvement on the DB side.  But even then, $150 a month is a bit too much for me to keep spending on something I'm not making any money from  I may try to cobble together something from the equipment I have available if I don't see much in the way of donations.

[flound1129]

FYI: You can buffer and compound multiple DB updates to the same table, if you use temp-tables to "hoard" the micro-updates. (Talking about shares that are slightly smaller and more realistic to less-loss.)

BTW if you have python code to do this I'd be interested in seeing it.  The current pool software already writes shares in blocks (I've upped the minimums on this recently which should help with performance)

[erk]

I have considered going all-scrypt, but we have some very dedicated TRC miners on the pool (it was one of the first coins I offered) and I do want to eventually include BTC once we have enough SHA256 hashrate.

As far as your db suggestions, I agree that there could be much improvement on the DB side.  But even then, $150 a month is a bit too much for me to keep spending on something I'm not making any money from  I may try to cobble together something from the equipment I have available if I don't see much in the way of donations.

The reason you have a loyal TRC following is the only other pool I know of is Coinotron, except for BTC, SHA256 coins are woefully serviced. The PPcoin devs haven't even made a port of p2pool for it, and how long has it been out?

Why not make a SHA256 and a scrypt version of the pool?

I would like to see TRC in rotation with PPC and BTC. That should attract a fair few from the new crop of smaller ASIC owners.

[Piotrsama]

06/06/2013 SXC will be removed from the frontend on or about June 8 at 00:00 UTC. Please withdraw your coins ASAP.

I'm trying to cash out to my SXC address, but it does nothing.
It doesn't even record my address on "Account details".
It's a bug or what?


EDIT: Solved by saving my address first, and later cashing out.
Maybe it helps somebody else.

[flound1129]

I have considered going all-scrypt, but we have some very dedicated TRC miners on the pool (it was one of the first coins I offered) and I do want to eventually include BTC once we have enough SHA256 hashrate.

As far as your db suggestions, I agree that there could be much improvement on the DB side.  But even then, $150 a month is a bit too much for me to keep spending on something I'm not making any money from  I may try to cobble together something from the equipment I have available if I don't see much in the way of donations.

The reason you have a loyal TRC following is the only other pool I know of is Coinotron, except for BTC, SHA256 coins are woefully serviced.

Why not make a SHA256 and a scrypt version of the pool?

I would like to see TRC in rotation with PPC and BTC. That should attract a fair few from the new crop of smaller ASIC owners.

Yes that is the eventual plan.  PPC needs some changes to both the mining software and the frontend, I believe.

We don't have a snowball's chance in hell of finding a BTC block in any reasonable amount of time until we have around 500GH on SHA256 though.

[roy7]

The reason you have a loyal TRC following is the only other pool I know of is Coinotron, except for BTC, SHA256 coins are woefully serviced.

I may be tiny, but my TRC pool does exist. Granted most network hash power seems to be at Coinotron.

[erk]

The reason you have a loyal TRC following is the only other pool I know of is Coinotron, except for BTC, SHA256 coins are woefully serviced.

Why not make a SHA256 and a scrypt version of the pool?

I would like to see TRC in rotation with PPC and BTC. That should attract a fair few from the new crop of smaller ASIC owners.

Yes that is the eventual plan.  PPC needs some changes to both the mining software and the frontend, I believe.

We don't have a snowball's chance in hell of finding a BTC block in any reasonable amount of time until we have around 500GH on SHA256 though.
[/quote] Quite true, BTC can take days, NMC is another SHA256 candidate, I don't mean merge minied, just mined by itself. I think that could be popular too.

[roy7]

Quite true, BTC can take days, NMC is another SHA256 candidate, I don't mean merge minied, just mined by itself. I think that could be popular too.

Wouldn't it be more profitable to merge mine NMC though?