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

[flound1129]

It's nothing obvious..  I think it may be a deadlock that's only happening intermittently.  Added a few debugging lines to the estimates script.

[Mahav]

really diggin' this pool. suppose this is a free bump then

[flound1129]

really diggin' this pool. suppose this is a free bump then

I'll take it!

[ranlo]

really diggin' this pool. suppose this is a free bump then

I'll take it!

How are donations going so far towards the increased costs of the server, btw?

[flound1129]

really diggin' this pool. suppose this is a free bump then

I'll take it!

How are donations going so far towards the increased costs of the server, btw?

Practically nonexistent!  Of course that may have something to do with the bug that was resetting everyone's donate percent to zero...  Which is fixed now!

[ranlo]

really diggin' this pool. suppose this is a free bump then

I'll take it!

How are donations going so far towards the increased costs of the server, btw?

Practically nonexistent!  Of course that may have something to do with the bug that was resetting everyone's donate percent to zero...  Which is fixed now!

Definitely possible. Also, your changes to optimize the site should hopefully reduce the costs by a bit (or a lot). Hopefully!

[flound1129]

really diggin' this pool. suppose this is a free bump then

I'll take it!

How are donations going so far towards the increased costs of the server, btw?

Practically nonexistent!  Of course that may have something to do with the bug that was resetting everyone's donate percent to zero...  Which is fixed now!

Definitely possible. Also, your changes to optimize the site should hopefully reduce the costs by a bit (or a lot). Hopefully!

Well, we're currently on an m1.medium instance at Amazon, which is around $95 a month not counting I/O, and to add more coins I'd really need to go to an m1.large or change the architecture a bit so that we're not using just one server.

[Zalfrin]

Maybe you could add a leader board for donation % if you get a chance. Try and get some competition going there.

I noticed my donation percent drop back to 0 at one point a couple days ago, but then I set it back again and it seemed to have stuck. Hopefully it's working like it should now, it claims to be on.

[flound1129]

It looks like if I move the shares tables to the instance store partition, I could save quite a bit on I/O costs.  If the hardware ever goes down, we would lose that data, however, I could easily backup the shares tables every hour to an EBS volume.  I think the risk of losing an hour's worth of shares is probably worth saving ~30-$40/month.

I've also identified an issue in the LTC shares table that, when fixed, should speed up the estimates calculation considerably.  I will be implementing this sometime tonight during a MNC mining window.

[flound1129]

Maybe you could add a leader board for donation % if you get a chance. Try and get some competition going there.

I noticed my donation percent drop back to 0 at one point a couple days ago, but then I set it back again and it seemed to have stuck. Hopefully it's working like it should now, it claims to be on.

It's on the short list of things to get done, along with detailed block payouts and personal payout history.  I'm at a client site for work today and tomorrow so it'll probably get done sometime this weekend.

[ranlo]

It looks like if I move the shares tables to the instance store partition, I could save quite a bit on I/O costs.  If the hardware ever goes down, we would lose that data, however, I could easily backup the shares tables every hour to an EBS volume.  I think the risk of losing an hour's worth of shares is probably worth saving ~30-$40/month.

I've also identified an issue in the LTC shares table that, when fixed, should speed up the estimates calculation considerably.  I will be implementing this sometime tonight during a MNC mining window.

I think it's worth the risk of an hour or so as well. Hopefully it won't come to that, but if it's a choice between the site and a small risk, or no site, it's an easy one to make, :p.

[Jazkal]

LuckyCoin or MegaCoin?

[jcrubino]

1) Can we have a client that changes hardware depending on the necessary algorithm?
           Device profile asic for sha-256
           Device profile gpu for scrypt+sha-256

[ranlo]

1) Can we have a client that changes hardware depending on the necessary algorithm?
           Device profile asic for sha-256
           Device profile gpu for scrypt+sha-256

I think someone would have to create that on their own, unless flounder (or someone else) created an app that would interface with the server and automatically close/launch the miner with the appropriate flags when it goes from scrypt to sha-256 coins.

[flound1129]

1) Can we have a client that changes hardware depending on the necessary algorithm?
           Device profile asic for sha-256
           Device profile gpu for scrypt+sha-256

I think someone would have to create that on their own, unless flounder (or someone else) created an app that would interface with the server and automatically close/launch the miner with the appropriate flags when it goes from scrypt to sha-256 coins.

Yeah, you'd need to ask the cgminer (or other mining app) developer to implement that.  I don't know if the software can autodetect the algorithm, although since stratum is http-based I could probably add an X-header tag or something.

[flound1129]

MySQL changes are complete.  Also, I reduced the cache times on many of the cached variables to 120 seconds.  Which means you should see things like hashrates, etc. update a lot faster.

[gnomicide]

1) Can we have a client that changes hardware depending on the necessary algorithm?
           Device profile asic for sha-256
           Device profile gpu for scrypt+sha-256

I think someone would have to create that on their own, unless flounder (or someone else) created an app that would interface with the server and automatically close/launch the miner with the appropriate flags when it goes from scrypt to sha-256 coins.

Yeah, you'd need to ask the cgminer (or other mining app) developer to implement that.  I don't know if the software can autodetect the algorithm, although since stratum is http-based I could probably add an X-header tag or something.

From what I've been able to gather, only the initial calls are http. If the call (getwork or getmininginfo?) returns a x-stratum header, cgminer then tries to connect to it. Subsequent calls are over that "stratum+tcp" protocol, and there are no headers, just the json values going back and forth.

I also believe at the miner level, you'd have to stop and restart the app, so something running locally could periodically query an API for "most profitable algorithm", then kill cgminer and reopen it with a different config file specified. On the pool end, you could just have a different port for each script, and the miner's config file points to the right port/algo.

[flound1129]

I'm compiling NVC on a t1.micro instance right now.

My plan is the following:  If things go well (i.e. if the micro instance is sufficient to handle the load of novacoind + stratum server, which it should be, in fact I may be able to put 2 coins on it) I plan to sell 10,000 WDC (~1 BTC, heh) from the pool reserves (about half our WDC reserves) to purchase a 3 year heavy utilization reserved instance ($100).  The recurring charges for the instance will then only be ~$10 per month.  ($3.75 for the server at $0.005 per hour, plus bandwidth and I/O, both of which should be fairly low.)

If things work out, I will probably move the other coins onto their own micro instances as I receive funds.  That frees up the current server to be a dedicated DB and would potentially allow me to move it outside of Amazon if I want to.

Anyone see a flaw in my thinking?

[flound1129]

1) Can we have a client that changes hardware depending on the necessary algorithm?
           Device profile asic for sha-256
           Device profile gpu for scrypt+sha-256

I think someone would have to create that on their own, unless flounder (or someone else) created an app that would interface with the server and automatically close/launch the miner with the appropriate flags when it goes from scrypt to sha-256 coins.

Yeah, you'd need to ask the cgminer (or other mining app) developer to implement that.  I don't know if the software can autodetect the algorithm, although since stratum is http-based I could probably add an X-header tag or something.

From what I've been able to gather, only the initial calls are http. If the call (getwork or getmininginfo?) returns a x-stratum header, cgminer then tries to connect to it. Subsequent calls are over that "stratum+tcp" protocol, and there are no headers, just the json values going back and forth.

I also believe at the miner level, you'd have to stop and restart the app, so something running locally could periodically query an API for "most profitable algorithm", then kill cgminer and reopen it with a different config file specified. On the pool end, you could just have a different port for each script, and the miner's config file points to the right port/algo.

Well, when the multiport switches, the miner has to make a new connection so I'd assume it will receive the X- headers again.  Not sure if the miner itself will look at these, though, or disregard them.

[CommanderVenus]

Am i not allowed to start mining Terracoin if Litecoin in the middle of the header on the page?

I have been trying everything to connect to stratum+tcp://multipool.in:3333 to mine TRC but i cant. I am using the correct username and password too.