[ANN] ptsweb.beeeeer.org - Protoshares mining sub-pool

[afr33sl4ve]

How do I update the github?

I try to run the command "git clone https://github.com/thbaumbach/ptsminer" and it returns that it already exists. I think I need to modify it to update as opposed to just simply creating the directory.

[montecristo83]

build notes:

boost library & YASM required for compilation!

Code: [Select]

sudo apt-get install yasm

Oh, I should have paid more attention to details like that. Thank you very much for pointing that out!

Was able to makefile and launch the miner. Just gotta disable AVX (currently reading about how to do it).

[GoldTiger69]

build notes:

boost library & YASM required for compilation!

Code: [Select]

sudo apt-get install yasm

[GoldTiger69]

I think you just need to install yasm

[montecristo83]

Having some issues with the latest linux x64 build on any of the 3 cloud services I've tried (DC, AWS & HP). Didn't have these kind of issues a few days ago (pretty sure I was using v0.6 or lower).

I've tried Debian 7 and almost all versions of Ubuntu to no avail.

After downloading and installing the github function, I clone ptsminer from github, then cd to its directory, and when I attempt to makefile, this error occur:

Code: [Select]

ubuntu@ip-xxx-xx-x-xx:~/ptsminer/src$  make -f makefile.unix
g++ -c -O2  -fpermissive -o obj/cpuid.o cpuid.c
yasm -f elf64 -o obj/sha512_avx.o intel/sha512_avx.asm
make: yasm: Command not found
make: *** [obj/sha512_avx.o] Error 127
If anyone can link me to v0.6 / v0.5 or share a solution to this issue, that'd be highly appreciated!

[xornof]

pull the latest github update 

Thanks, new compile went without a hitch, looks good so far.
I had the counter showing <10c/m this morning, i initially sobbed , and then i checked my overnight totals and they were up 10% from the night before, woot...

Thanks for all your efforts!

[cjrhoades]

Same computer, but different order of the images; the last one should be the first one. Just saying.

You'd be correct. Same computer, different times. The 9ish hour long one actually came first. I started it up with similar performance to whats in the first screen shot, went to bed, and woke up with what you see there and took a screenshot. Later on I restarted it for additional screenshots to show the decreasing performance issue. Sorry if that happens to look shady.

it's not slowing down (or at least it shouldn't from a coding perspective)
but the counter had a bug, showing wrong values after a connection loss / reconnect

I see. Thanks for fixing it.

[xolokram]

it's not slowing down (or at least it shouldn't from a coding perspective)
but the counter had a bug, showing wrong values after a connection loss / reconnect
pull the latest github update 

i'm sorry for the inconvencience

btw. last pool software restart (crash?) was 6 hours ago!

[GoldTiger69]

Thank you for your hard work, xolo. Nevertheless, I've been experienceing the same problem with the 0.7a in Ubuntu server 13.10. After several hours, the miner solwdown by 50% or so. When you re-start the miner, it comes to normal again.

[xolokram]

oh, thanks for pointing that out @GoldTiger69
this makes your statement look weird @cjrhoades

[GoldTiger69]

That data must be from different computers. The clocks on the images say's so 

OR

Same computer, but different order of the images; the last one should be the first one. Just saying.

[xolokram]

so the "decreasing" performance is was just a stupid bug in the counter or to be more precise, like i mentioned before, the col/min & sh/min counter was not reset properly on a connection loss
i fixed that with the last github commit, i'll update the windows binary in a few minutes


if your miner states ~1400 shares and the last blocks contains less than that, then the submission didnt reach the server

/edit1:
OR

the only thing i can image of is that the pool software crashed sending some of the shares of the current round into nirvana (for all users)
let me check the time of the last restart of the pool software
(this should also be the cause of your connection loss & buggy col/min+sh/min reset)

- xolokram

[cjrhoades]

@cjrhoades / afr33sl4ve:
this is getting a little confusing, so you're using v0.5 and v0.7 ? both with the same problem.
and both of you encounter the "collision found" but -no- "submitted share -> <whatever>" line in the miner output?

No. The shares are being submitted fine. Here's my concern.

After a fresh start:


After ~40 minutes:


After ~9 hours:


Notice the big dip in performance over time? The longer the miner runs, the slower it gets. Closing and restarting it brings it back up to maximum speed.

also please be aware, that your most recent shares are currently not shown in the user stats, as these are bound to found blocks, and the last found blocks is 2.5 hours old

- xolokram

Ok, here's what I know. I've put in 1400-1500 shares so far. All the blocks shown in the user stats add up to ~900 shares. If it hasn't been updated for 2.5 hours, then during that time I would have had to put in ~550 shares and been doing around 3.6 sh/m to have that many missing. But I've been doing far less than that. So either its taking longer than that to update or some shares aren't being counted. Or my math is wrong in which case please ignore this. What I do know is that every time a collision is found I see "[MASTER] submitted share -> SHARE" so I know they're getting to you.

[xolokram]

I'm seeing, though, lower hashrates as time goes on, with both 0.5a and 0.7a.

this can happen if your miner loses the connection to the server due to a bug in the col/min counter (v0.7)
i'll commit a fix in a few minutes

v0.5 is strange, as this version doesnt have the mentioned bug. maybe your cpu is becomming hot and throttles down? can you check that please.

[afr33sl4ve]

No, no. I'm getting rejections and collisions, however, they're all being submitted. I'm not having the no submissions issue.

I'm seeing, though, lower hashrates as time goes on, with both 0.5a and 0.7a.