[ANN] profit switching auto-exchanging pool - www.middlecoin.com

[MiMiMiner]

Does anyone know why I would be getting roughly 150-250mhsh slower (per machine) on middlecoin than I do from any other pool?

I would like to mine here, but with three machines, this is making a difference of about 600 mhsh that appears to be going nowhere.

If I had a reasonable explanation for this, I'd probably choose to continue mining here, as it has generally been good to me (THANKS, H2O!).

I have been noticing this for the last five days or so.

cause it's switching coins ever so often.

But, does the switching make up the low hashrates?

[1715323041]

1715323041

[1715323041]

1715323041

[1715323041]

1715323041

[Test User]

But, does the switching make up the low hashrates?

Every time the coin switches, cgminer will pause for a few seconds and discard the work it was doing on the old coin, so those hashes don't get counted.

There is also the issue that the stratum mining protocol isn't set up for changing coins on the same pool. The pool has to reset the connection, and cgminer sees this as a network problem, and will stop mining, or change to another pool.

The problem is that middlecoin is changing coin every 30-45 seconds, so when cgminer goes idle for 5 seconds after a coin change, it makes a big overall impact.



You can see this happening in this monitor screenshot from my mines. The orange hash rate charts are middlecoin, and you can see the pauses every time the coin changes. For comparison the blue and red charts are different pools, and you can see the highly consistent hashrates. You can also see how the coin change affects all the mines simultaneously.

[spip]

Guys I wrote a quick and dirty calculator that estimates profit of this pool and thought I'd share.

It extracts h2o's profit data and extrapolates that trend. Obviously we can't predict the future trend, but this should get you a short term ballpark.

https://docs.google.com/spreadsheet/ccc?key=0Ahtw8TH0C2qIdEJSYTRmZEpZRXBVRzlGQTg0QWdmRkE&usp=drive_web#gid=0

[glon]

Could H20 remove all the addresses without any hashing power? Or at least remove the ones that haven't mined on middlecoin for a certain amount of time... It takes forever to load the main page and that's because there are so many addresses to load. More than half of the addresses on the page do not have any hashing power.

Use this if you want been working on it for a little while and everything seems to be ok minus updates are not updating for the past couple of hours as other have shown but you can go directly to your information by using the w={WalletID} in the url

http://www.bfpcoin.com

What do you mean by "updates are not updating"? My miners are currently mining at 2.4MH/s but your tool has been showing 1.4 MH/s for me ever since I tried it some 30 minutes ago. Is that what you mean? That this 1.4MH/s is some stale number from a couple of hours ago? Thanks!

[neofatalist]

Mining for 2 days and my address wont show on main page. Sorry if I seem spamming but my posts wont show.
1JGJhJK9bB6ZRxpBtiSUt1NtAXnATpuA5Q

[CrimsonPermanentAssurance]

You can see what time (in UTC time) the last update occurred to determine if it's stale or not. It shows what time it last updated in the chart.

[CrimsonPermanentAssurance]

But, does the switching make up the low hashrates?

Every time the coin switches, cgminer will pause for a few seconds and discard the work it was doing on the old coin, so those hashes don't get counted.

There is also the issue that the stratum mining protocol isn't set up for changing coins on the same pool. The pool has to reset the connection, and cgminer sees this as a network problem, and will stop mining, or change to another pool.

The problem is that middlecoin is changing coin every 30-45 seconds, so when cgminer goes idle for 5 seconds after a coin change, it makes a big overall impact.



You can see this happening in this monitor screenshot from my mines. The orange hash rate charts are middlecoin, and you can see the pauses every time the coin changes. For comparison the blue and red charts are different pools, and you can see the highly consistent hashrates. You can also see how the coin change affects all the mines simultaneously.

So, I assume that same issue would also cause invalid shares to double from what they were doing prior to this problem. This is what I seem to be be experiencing. Is that the case? I have two machines, each with four 7950s that are hashing at about 2.39 MH/s each with a whopping 7%plus being rejected. Those numbers were closer to 3% rejected and 2.54 MH/s each six days ago. No changes have been made to the conf files in that time. Is there a change we should be making to the conf files, or is this what everyone is experiencing and considering to be acceptable? I am on port 3333 of the US server, but from my own experiments, it seems no different on port 8080 of the US server.

[Test User]

I don't know if coin switching alone will account for the rejects. From what I have seen, when the coin changes, cgminer simply discards the work it was doing, and doesn't count it as a reject.

Rejects tend to occur when the work is received too late. I find that this happens most often when mining very "fast" coins. Mining litecoin, which has a 2.5 minute block time, I get about 3% rejects, mining DOGE with a 1 minute block, I get about 4%, for something like casinocoin with a 30 sec block, I get about 7%, and for something like fastcoin, I get over 20% rejects. A slow or laggy internet connection (e.g. a 3G connection, which frequently has a ping time in the 2000 ms range) can make this problem a lot worse.

The same problem can occur when a huge pool switches onto a small coin, and the block rates go way out of spec before the difficulty readjusts.

I haven't noticed middlecoin mining any very fast or very small coins. At the moment its flip-flopping between litecoin, dogecoin and megacoin every few seconds, but I'm not getting more rejects that I normally get. However, my miners keep failing over onto the EU server several times per hour, presumably because there are too many coin switches and disconnects.

[envy2010]

Mining for 2 days and my address wont show on main page. Sorry if I seem spamming but my posts wont show.
1JGJhJK9bB6ZRxpBtiSUt1NtAXnATpuA5Q

Can you post your miner config file/startup command line?

[CrimsonPermanentAssurance]

I don't know if coin switching alone will account for the rejects. From what I have seen, when the coin changes, cgminer simply discards the work it was doing, and doesn't count it as a reject.

Rejects tend to occur when the work is received too late. I find that this happens most often when mining very "fast" coins. Mining litecoin, which has a 2.5 minute block time, I get about 3% rejects, mining DOGE with a 1 minute block, I get about 4%, for something like casinocoin with a 30 sec block, I get about 7%, and for something like fastcoin, I get over 20% rejects. A slow or laggy internet connection (e.g. a 3G connection, which frequently has a ping time in the 2000 ms range) can make this problem a lot worse.

The same problem can occur when a huge pool switches onto a small coin, and the block rates go way out of spec before the difficulty readjusts.

I haven't noticed middlecoin mining any very fast or very small coins. At the moment its flip-flopping between litecoin, dogecoin and megacoin every few seconds, but I'm not getting more rejects that I normally get. However, my miners keep failing over onto the EU server several times per hour, presumably because there are too many coin switches and disconnects.

Thank you for the quick answer, Test User. It just seems strange all my numbers across three machines are so out of whack from where they were six days ago (and only on middle coin servers). H20, forgive me if you have already addressed this problem previously, but is this an issue your server upgrades are going to address?

[Tripjammer]

I need a scrypt asic!!! LOL!

Damn peeps got 160,000 kh/s? Daaam!

I would be happy with 25,000 kh/s....

Looks like H20 is coming through for us today...payouts are starting to look good....

[hodger]

The EU server isn't very profitable at the moment. It's constantly mining DOGE.

If you have less than 250 ms ping to the US server, I suggest mining there, even though you get disconnects on coin switches.

I'll be upgrading the EU server to switch coins just like the US server within the next couple of days.

I was wondering what sort of ping response cutover figure to aim at.  I'm getting sub-200 on US, so over I go (for now at least).

Keep up the good work guys.

[coinGeek]

Woo Hoo!

The unexchanged is exchanging. The question is, will it hit the MOON for the first time?

[zSprawl]

Woo Hoo!

The unexchanged is exchanging. The question is, will it hit the MOON for the first time?

haha yeah I just saw my red line shoot upwards

[phzi]

A tip that might be of interest to some others:  Using load-balancing in cgminer seems to be smoothing out the frequent coin-switch disconnects on the US server:

Config example:

"pools" : [
        {
                "quota" : "8;stratum+tcp://middlecoin.com:3333",
                "user" : "",
                "pass" : "x"
        },
        {
                "url" : "stratum+tcp://eu.middlecoin.com:3333",
                "user" : "",
                "pass" : "x"
        }
],
"load-balance" : true

This will get the majority of work (8/9) from the US server, and the other 1/9th from the EU server.  When a disconnect happens on pool 0, you will usually have some work from pool 1 to switch to right away, so you lose a lot less hashing time vs sitting idle.  It isn't perfect - still losing a bit of hash rate on switches, but it doesn't seem anywhere near as bad as without.

[bal3wolf]

A tip that might be of interest to some others:  Using load-balancing in cgminer seems to be smoothing out the frequent coin-switch disconnects on the US server:

Config example:

"pools" : [
        {
                "quota" : "8;stratum+tcp://middlecoin.com:3333",
                "user" : "",
                "pass" : "x"
        },
        {
                "url" : "stratum+tcp://eu.middlecoin.com:3333",
                "user" : "",
                "pass" : "x"
        }
],
"load-balance" : true

This will get the majority of work (8/9) from the US server, and the other 1/9th from the EU server.  When a disconnect happens on pool 0, you will usually have some work from pool 1 to switch to right away, so you lose a lot less hashing time vs sitting idle.  It isn't perfect - still losing a bit of hash rate on switches, but it doesn't seem anywhere near as bad as without.

i tested that out thru out the day and still get some getwork failures on us but not as bad with the recompiled cgminer and load balance im gettng 1 getwork failure every 3 mins right now so far better then it used to do.

[Gamer67]

Not complaining just saying.

Today was a lower payout than expected. I guess those who stayed on the US server will be getting better payouts if the chart is to be believed. I attempted to switch to the US server yesterday (I am based in the US) but even with the recompiled miner the connection interrupts and constant switching made my hashrate plummet so I thought it would be best to stick with the EU server.

[bal3wolf]

Not complaining just saying.

Today was a lower payout than expected. I guess those who stayed on the US server will be getting better payouts if the chart is to be believed. I attempted to switch to the US server yesterday (I am based in the US) but even with the recompiled miner the connection interrupts and constant switching made my hashrate plummet so I thought it would be best to stick with the EU server.

i bouced between us and eu server last 2 days yesterday i made 31usd today i made 25usd with 2.3mhash.

[envy2010]

Not complaining just saying.

Today was a lower payout than expected. I guess those who stayed on the US server will be getting better payouts if the chart is to be believed. I attempted to switch to the US server yesterday (I am based in the US) but even with the recompiled miner the connection interrupts and constant switching made my hashrate plummet so I thought it would be best to stick with the EU server.

Doge is plummeting relative to BTC, so since that's what the EU server is mining, it won't earn you as much. The EU server is much is more stable though.

[Gamer67]

Not complaining just saying.

Today was a lower payout than expected. I guess those who stayed on the US server will be getting better payouts if the chart is to be believed. I attempted to switch to the US server yesterday (I am based in the US) but even with the recompiled miner the connection interrupts and constant switching made my hashrate plummet so I thought it would be best to stick with the EU server.

Doge is plummeting relative to BTC, so since that's what the EU server is mining, it won't earn you as much. The EU server is much is more stable though.

Fair enough. I'll take the good with the bad. Payout was better than if I mined LTC so it's all good.