 Show Posts
|
|
Pages: [1] 2 3 »
|
Wow. Many thanks for this. I didnt have time to poke at it very much but this makes it easier I had read about the challenge question but didnt quite get where the cf cookie was coming from when I was looking at my packet traces |
|
|
|
so I may have missed this in the many pages here, but have we ever gotten an API?
The whole cloudflare/ddos protection stuff is nice to keep the pool up, but incredibly annoying for both checking hte pages and trying to write data scrapers to track my performance.
If there is no API (still) any one have any suggestions on how to scrape data from their miner pages? I cant seem to get around the cloudflare junk
I am having the same problem in trying to "data scrape" as you call it. I read on the CloudFlare site that their protection does not prevent web crawlers from scanning the site (so I am confident we can find a way to do it), and I'm thinking what I will try next is to write code to accept their cookie(s). Of course, my guess is worthless until I have it running. What I unsuccessfully tried in php was to file_get_contents, then do a javascript setTimeout of 10 seconds, after which I do a window.location.reload. hmmm, well i guess if they say web crawlers can scan it, there has to be a way. what's the link where you read this? I went the python route a few weeks back, but got no where. After reading your post, i fired wireshark back up and it looks like there's a hidden form that needs some special key to return in order to let us through ... so I also think we'll need some code to accept their cookie/whatever and respond back similar to how a browser would. still kind of pain to grab some data  update: ok so i did some digging. got this from the cloudflare site: Javascript and cookies are required for the tests, and to record the fact that the tests were correctly passed. The page which your visitors see when in IUAM can be fully customized to reflect your branding. I'm Under Attack mode does not block search engine crawlers or your existing CloudFlare whitelist. My guess is they whitelisted known and allowed search engine crawlers which is why those bots work where as ours dont. Add to it that it looks like Terk opted to put the domain in a permanent I'm Under Attack Mode and we are where we are. So pretty much we need to mimic the expected behavior if we want anything to work ... lovely |
|
|
|
|
so I may have missed this in the many pages here, but have we ever gotten an API?
The whole cloudflare/ddos protection stuff is nice to keep the pool up, but incredibly annoying for both checking hte pages and trying to write data scrapers to track my performance.
If there is no API (still) any one have any suggestions on how to scrape data from their miner pages? I cant seem to get around the cloudflare junk
|
|
|
|
Any chance of adding in an API to access our user stats? I did some searching, and it doesn't look like there is one currently available.
^ This, please. Since I infrequently update the stats and balance pages I keep getting cloudflare "checking your browser" page. What's worse, my simple data scraping scripts broke b/c they dont respond like a web browser  |
|
|
|
|
has anyone mining on this pool been affected by the whole client redirect hijacking going through some of the other multipools?
|
|
|
|
|
Terk,
I've been scraping my miner's data off the balances site every 10-15 minutes over the past few days.
Today, I've been getting 503 service unavailable messages.
I'm guessing it's a rate limiting/ddos protection feature of the hosting service. Is there a good refresh frequency I can use to avoid getting the 503s?
|
|
|
|
Updates for today:
We pushed out an update that should improve our hashrate headroom dramatically. If you were previously on the EU server having having issues with not seeing your full hashrate, that should be completely resolved now. Sorry for the inconvenience, and please let us know if you try again and still have issues. We're currently processing ~3,000 shares per minute, and should have no problem scaling up past 50,000+ (fingers crossed).
We restored the EU VARDIFF to the previous target of 1 share / 25 seconds. It had been increased to 1/35 temporarily to reduce strain on the system.
We've also dropped the min VARDIFF target to 64, and it now moves in smaller increments (also 64). That should help us to support those of you with smaller rigs.
Tomorrow we plan on adding the coins we promised for this week, and removing some dead coins. If all goes well on the stability/profitability side, we'll also have some time to put in to the UI and API.
Hey Coinshift, Any possibility to add a running total of how much BTC the pool has paid out to the api page http://coinshift.com/api/stats/performance/?thx |
|
|
|
poolwafle : do i need to tweak my expiry setting to avoid stales ?
do they mean any kind of loss when if they are accepted ?
on another topic
with all the extra server capacity : what u think about seting up a scrypt-n based "altpool" on a different port ?
or the current sever already capable to handle scrypt-n ?
is there actually ANY scrypt-n based multipool around ?
could it be a possible direction for WP to evolve , or pool mining those would simply destroy that market ?
or its just way to early to work on or even to think about the implementation ?
yeah i suck at english but i think u might get my point
There is another multipool that has a scrypt n port. I think it closed registration. I am not involved with it(but do mine there at times) but I'll also not name it in this thread as I dont think it is right to advertise a diff pool in another pool's thread. |
|
|
|
some observations after another 24 hrs mining here. I had moved off to another pool previously, but came back and had a balance of 0, so this is fairly accurate. right now the api is returning "value_per_mh_24h": "0.00576" but my actual calculations for btc/mhash/day is roughly .0064 and .0063 across 2 different rigs I've also noticed we've been in a decline in profits for a few hours In the meantime, I do have one request. Would it be possible for the api page: http://coinshift.com/api/stats/performance/ to also include total payouts in btc? I'm not looking for a "for the day" total, I'm just looking for a "since the beginning" or "from this day forward". Essentially something that will not/never be set back to 0. I'm writing some scripts to pull out pool vs rig performance, so it would be helpful if there was a way to determine much btc was paid out in total. as it stands now, I can tell how much MY last payment(s) were (all time payouts on miner page, payout_sum on api page), but I have no idea how much in total was paid out to the pool. As far as I can tell the "Current Pool Balance" from the stats Page/value_24h from the api will get zeroed out every time there is a payout. -spuds |
|
|
|
|
figured I would chime back in.
reran the #s at 24 hours and i ended up with .00610 btc/mhash/day ... which was exactly what the stats page was listing at the time.
guess i just hopped on at the wrong time and missed the higher earning hours that drove the trailing 24 hrs # so high.
either way, i'll leave my rigs here for a bit and see how it goes in a few days.
|
|
|
|
|
any one else not really seeing the similar performance #s to the ones listed for trailing 24 hours?
i'm willing to give this pool some more time, but as of now (~17.5 hours in) I'm only getting about .0057/.0058 btc/mhash/day across 2 different rigs and not the roughly .007 as listed on the stats page. I included payouts, confirmed exchanged and estimated unexchanged in my calculations
I'll check back in a few hours once my full day comes around. I really want to believe in this pool since it is doing what a lot of the other pools should have been doing in the first place.
Would it be possible to represent profitability with additional details beyond trailing 24 hours only? Maybe break it down hourly for the past N hours or something? Does the API support that time granularity or is it only on 24 hr ranges?
|
|
|
|
richmke, that speaks bad for the configuration on the miners, don't you have a backup pool? Like EU primary, useast as backup1, etc? This way when any falls it just switches till it is over.
I start cgminer with a batch file. How can i setup a backup pool in the batch file? add the following at the end of your existing batch script --failover-only -o stratum+tcp://blah -u blah -p blah basically the basic cgminer command without the --scrypt flag |
|
|
|
Risers can now be ordered on Amazon! Soon to be fulfilled by Amazon, as soon as I can get enough here to cover my own customers and ship several thousand to the amazon warehouse.
any idea what you're planning on setting the price on amazon at? |
|
|
|
Anyhoot, that motherboard is an MSI motherboard that was made specifically for Dell. It has 3x x16 slots, and 2x x1 slots. You can find them on ebay for like $50 to $75. "Dell 730x motherboard" Make sure they are not "Dell 730" <- that "X" in the 730X means X58 INtel chipset. A "non-X" means NForce LGA775 socket!
A virgin copy of LTCRabbit freshed minted onto an USB, booted, and BAM - all 5 GPUs mining away (no tweaks yet).
I really want to believe this mobo can be had for 50 to 75 on ebay, but a search for sold or completed listing using "Dell 730x motherboard" returns items that were sold for $120-$300 going back to November. Anything in the $50 range is nothing more than a mobo tray. Like I said, I WANT to believe this to be true, as I would gobble up some boards. But it just doesn't seem to be the case on ebay right now... how long ago did you grab the mobo? |
|
|
|
I won't buy these until I can get more than 3 at a time.. I refuse to buy if I'm forced into the highest price.
That's fine, I'm happy to mine at the lower difficulty while you wait  I ordered 3 yesterday, but I see that they are now sold out. I also bought ten of the 5x gc3355 round miners (300kh/s ea.) and I am looking for larger quantities of basically anything with the gc3355 chip in it. where did you order the round miners? if i had to guess, i would say: https://bitcointalk.org/index.php?topic=421921.0same seller has them up on aliexpress too |
|
|
|
With those ping times, I'd definitely move one to useast and one to eu and start testing. My times to eu are 3x that of the us servers, so I've never bothered.
well somewhere along the line, useast dropped from 100 ms to 10 ms latency. pointed 1 rig @ useast and 1 rig @ eu. will report back in a few days on my findings minor update: eu is reporting 200-300 less WU after running for about an hour. previously, uswest was a consistent 1.3. so, for me eu is definitely no good. moving both over to us east asap minor update2: anyone else with 2 mining rigs @ the same server observe that sometimes the cgminers don't have the same difficulty??? Sometimes, i see one of my miners takes a few minutes to get the network difficulty change. So much infact, that i just saw one of my rigs not get the network diff change at all. It went something like 488->11->11->10 while the other one happily sat at 488 and both were getting accepted shares at the different difficulties  other times they both get the diff change within seconds. it just seems kind of odd |
|
|
|
Yes, just add "--failover-only" to your command line arguments.
Why "ride it out"? I'm assuming you have historical stats to compare. Just create a new address and switch, then compare results after a few days.
that's a good point. while my 2nd rig only has a day and a half on middlecoin, my other rig has enough historical data to show me i'm at 0.008/0.009 btc/mh/day. i wonder if it's worth testing eu for me as well since I have about the same latency to all 3 Pinging uswest.middlecoin.com [54.214.242.184] with 32 bytes of data: Reply from 54.214.242.184: bytes=32 time=97ms TTL=245 Reply from 54.214.242.184: bytes=32 time=97ms TTL=245 Reply from 54.214.242.184: bytes=32 time=96ms TTL=245 Reply from 54.214.242.184: bytes=32 time=96ms TTL=245 Pinging useast.middlecoin.com [54.197.251.210] with 32 bytes of data: Reply from 54.197.251.210: bytes=32 time=105ms TTL=251 Reply from 54.197.251.210: bytes=32 time=103ms TTL=251 Reply from 54.197.251.210: bytes=32 time=105ms TTL=251 Reply from 54.197.251.210: bytes=32 time=96ms TTL=251 Pinging eu.middlecoin.com [54.194.173.83] with 32 bytes of data: Reply from 54.194.173.83: bytes=32 time=94ms TTL=250 Reply from 54.194.173.83: bytes=32 time=98ms TTL=250 Reply from 54.194.173.83: bytes=32 time=95ms TTL=250 Reply from 54.194.173.83: bytes=32 time=95ms TTL=250 |
|
|
|
My imprecise results since the last payout...two different addresses used.
middlecoin = 0.005286 BTC/Mh/s (about 2.23 Mh/s)
useast = 0.005322 BTC/Mh/s (about 2.03 Mh/s)
I'll continue to monitor. Periodic checks throughout the day showed that both pools were mining the same network.
I assume you mean that was your average between the last TWO payouts? That seems awfully low. I'm seeing > .01 BTC / MH / day. here's my results using middlecoin.com: rig 1 (1.4Mh) = .0123 BTC/day rig 2 (2.8Mh) = .027 BTC/day (need more days for better analysis) but basically this comes down to .008 BTC/MH/day or .009 BTC/MH/day if what you guys are saying is right, i'm losing a fair amount per day by not using useast that's just great. guess i'll ride out uswest till tomorrows payout and then swap over to useast and do some more testing, though I'm not sure now is the best time to do comparo tests given the fact that the graphs are totally fubared for failover only, we just append the following to our scripts (or the equivalent in the config)? --failover-only -o stratum+tcp://middlecoin.com:3333 -u <your bitcoin address> -p x |
|
|
|
useast has much better payouts than uswest. I assumed everybody knew that.
so uswest == middlecoin.com and useast is still beta and they are mining differently. hm... i've been trying to follow the differences b/w the servers but it is kind of hard to keep up with the all the extra crap going on in the thread. I was aware that middlecoin and eu.middlecoin were different for a while, is that still true? Also with useast still in beta, does it still go down occasionaly? what's the best approach, use useast with middlecoin.com as --failover-only so that you dont suffer any downtime when/if useast goes down? now that my second rig is pointed here, i was planning on doing some tests to the various servers. i'm about 100ms latency to useast, uswest and eu. also, for what it's worth i'm seeing slightly different results for my btc/MH/day with 2 different miners pointed at middlecoin.com. need a few more days i think before i make my own conclusions on the whole new vs old address profitability debate |
|
|
|
|
