GPUMAX | The Bitcoin Mining Marketplace

[jamesg]

They are manually approved and Pirate is sleeping. 

I can understand manually approving purchases.
What I find harder to understand is that queued (ie approved) purchases are still not processed while he is asleep.
Gpumax is what, 5 months old now? Its not criticism or that I dont understand beta status, but I just dont understand whats so different about processing leased shares compared to passthrough.

Until they have failover for purchases in the production environment they will not be running purchases at night.

[P4man]

Until they have failover for purchases in the production environment they will not be running purchases at night.

Well, Im not privy to how gpumax works internally, so I may be missing something, but clearly they already have a robust form of failover: the private pools.
If something goes wrong processing leased shares, just direct the miners back to their private pools. Thats not worse than what it is now, in fact, its the exact same thing - except if it there is no problem, you can process leases even if pirate sleeps .

Im sure there is a good reason why this isnt done, I just wonder what it is.

[boozer]

Until they have failover for purchases in the production environment they will not be running purchases at night.

Well, Im not privy to how gpumax works internally, so I may be missing something, but clearly they already have a robust form of failover: the private pools.
If something goes wrong processing leased shares, just direct the miners back to their private pools. Thats not worse than what it is now, in fact, its the exact same thing - except if it there is no problem, you can process leases even if pirate sleeps .

Im sure there is a good reason why this isnt done, I just wonder what it is.

My understanding is they want to monitor all the purchases because a plethora of things can go wrong that does not cause failover to private pool as of yet... like if the purchase goes to a pool that can't handle the load, then we get lots of stales/rejects, so he wants to monitor that so our mining time isn't wasted.... here's what Pirate said a few months ago:

It's not really a problem as much as it is people buying shares on pools that can't handle the load which we are still fine turning to protect our miners.  Updates coming soon will address this and make buyers pay for rejects on pools that have issues so the miners are not penalized for it.

EDIT: We get a lot of pools ops testing their system which is not something miners should have to pay for with rejects and stales.

[P4man]

My understanding is they want to monitor all the purchases because a plethora of things can go wrong that does not cause failover to private pool as of yet... like if the purchase goes to a pool that can't handle the load, then we get lots of stales/rejects, so he wants to monitor that so our mining time isn't wasted....

Well, monitoring stales isnt that hard. Im sure a 5 line script can do that (and almost certainly already does). The same goes for any other criterium I can think off. Why do you want a human to look at those numbers all day long? We arent expecting pirate, or anyone else for that matter, to stare at stats 18 hours per day, are we? Thats why we program things .

Also you could put very conservative tresholds in those scripts that cause a fallback to private pools, its not like that is a net loss to anyone compared to what it is now.  The current system was understandable months ago, but at this point I find it a bit odd.

[jamesg]

The current system was understandable months ago, but at this point I find it a bit odd.

Please go somewhere else then.

[P4man]

The current system was understandable months ago, but at this point I find it a bit odd.

Please go somewhere else then.

Oh common, there is no need for that kind of response. Its a genuine question. If you know the obvious answer, why not tell us?

[jamesg]

Oh common, there is no need for that kind of response. Its a genuine question. If you know the obvious answer, why not tell us?

The obvious answer is that zux0r is working hard to make the system the best it can be.

It is only your reaction to the situation which suggests that fine tuning software that does 15-20k db writes a second is no big deal to complete. While the interface may be simple (or look simple) i can only imagine the complexities involved in make GPUMAX work like clock work 24/7/365.

Have patience.

[P4man]

It is only your reaction to the situation which suggests that fine tuning software that does 15-20k db writes a second is no big deal to complete.

No, and thats the point. I never implied building something like gpumax was trivial, please dont make it sound like I did.
But monitoring some vital stats like stales and throughput through scripts rather than having a human looking at something 16 hr per day (and that something can only be generated by.. such scripts) would indeed seem trivial by comparison.

[notme]

It is only your reaction to the situation which suggests that fine tuning software that does 15-20k db writes a second is no big deal to complete.

No, and thats the point. I never implied building something like gpumax was trivial, please dont make it sound like I did.
But monitoring some vital stats like stales and throughput through scripts rather than having a human looking at something 16 hr per day (and that something can only be generated by.. such scripts) would indeed seem trivial by comparison.

So on top of those 15-20k writes per second, you want to add an extra check to each one to tell if the share is stale.  If it's so trivial, why don't you buy Pirate an extra server to handle the added load?

[P4man]

So on top of those 15-20k writes per second, you want to add an extra check to each one to tell if the share is stale.  If it's so trivial, why don't you buy Pirate an extra server to handle the added load?

What are you saying, pirate is checking 20K shares per second for stales by hand when he is awake?
Let me rephrase the question: what is it you can check manually by looking at a screen, that you can not monitor automatically relatively easily ?

[notme]

So on top of those 15-20k writes per second, you want to add an extra check to each one to tell if the share is stale.  If it's so trivial, why don't you buy Pirate an extra server to handle the added load?

What are you saying, pirate is checking 20K shares per second for stales by hand when he is awake?

Yes.  He is superman.

[kitich]

~2h

[04:02:44] Couldn't connect to server, retrying...
[04:03:04] Couldn't connect to server, retrying...
[04:03:24] Couldn't connect to server, retrying...
[04:03:44] Couldn't connect to server, retrying...

something wrong?

[bulanula]

Hello.

Purchase 1 has been created and is pending payment.
about 9 hours ago

~50 network confirmation, when my purchase will be ok? )

They are manually approved and Pirate is sleeping.  It is 2AM in his timezone.

What timezone would that be ?

Thanks !

[notmove]

~2h

[04:02:44] Couldn't connect to server, retrying...
[04:03:04] Couldn't connect to server, retrying...
[04:03:24] Couldn't connect to server, retrying...
[04:03:44] Couldn't connect to server, retrying...

something wrong?

Same here for a Couple of days.

[DILLIGAF]

Hello.

Purchase 1 has been created and is pending payment.
about 9 hours ago

~50 network confirmation, when my purchase will be ok? )

They are manually approved and Pirate is sleeping.  It is 2AM in his timezone.

What timezone would that be ?

Thanks !

Central Daylight Saving Time by now in the US.

[GenTarkin]

Is the price really almost .00006??!?! whats driving it up!?

[jamesg]

Is the price really almost .00006??!?! whats driving it up!?

Looks to me to be 0.00005341.

[bitlane]

Is the price really almost .00006??!?! whats driving it up!?

The GPUMAX PPS price jumps have been all over the board.
I just got tired of 'chasing' the current rate around, so I set mine low enough not to have to worry and high enough to make it worth my while.

Just a shot in the dark, but I have a hard time believing that it's large miners that are causing the price swings.
With the amount of time to edit each of their workers, I just don't see it.

So, as a small FYI, here's what a 5GH/s setup looks like (rough averages....obviously YMMV)

The 'average' 5GH/s Setup.
--------------------------
- 70 shares/minute
- 4,200 shares/hour (60 minutes)
- 100,800 shares/day (24 hours)

Now, estimate that GPUMAX makes up about 50% of your daily mining shares/income.

100,800 shares/day (24 hours) = 50,400 shares/12 hours average.

Using the above 12 hour share average (for a 5GH/s Miner), here's the difference between a few GPUMAX PPS rates:
0.000050 BTC/share = 2.520 BTC per half day average (5GH/s Miner).
0.000045 BTC/share = 2.268 BTC per half day average (5GH/s Miner).
0.000040 BTC/share = 2.016 BTC per half day average (5GH/s Miner).

The above figures assume near-perfect conditions, which are difficult to achive in any venture.

Currently, the rates (from what I have seen) have been anywhere from .00004 to .00005 NET (add 0.000005 approx. for GROSS rate).

Let's further reduce this down to support a 1GH/s average Miner:
0.000050 BTC/share = 0.5040 BTC per half day average (1GH/s Miner).
0.000045 BTC/share = 0.4536 BTC per half day average (1GH/s Miner).
0.000040 BTC/share = 0.4032 BTC per half day average (1GH/s Miner).


For the TINY gain in BTC for anyone under 5GH/s......Is it really worth all the fucking around ?

As others have pleaded: Set a price that is fair and forget it. Quit chasing pennies.

Cheers,
bitlane.

[notme]

Just a shot in the dark, but I have a hard time believing that it's large miners that are causing the price swings.
With the amount of time to edit each of their workers, I just don't see it.

I'm pretty sure you can point multiple miners at the same "worker".  Also, large miners have software to aggregate their miners internally and only make one connection the the pool.

[cablepair]

heheh damn I have been missing out

I set mine at .00004 weeks ago and have been so busy I have not paid any attention to it

excuse my ignorance but gigavps: how are you determining the exact amount?

thanks