Topic: [ANN] dstm's ZCash / Equihash Nvidia Miner v0.6.2 (Linux / Windows) - page 31. (Read 224961 times)

This is a very strange issue, in Windows 10, the 0.6 miner is stable but once you randomly click inside the cmd window, it will just freeze (as in not outputting any new lines), the GPUs clocks are reduced to around 1500MHz (my cards are 1080Tis). I have to close the window and start the miner again. I can reproduce the issue at any time and happens in all of my rigs, so the issue must be not in my system.

Not sure why this is happening, it is not a critical issue but dstm you might want to look at it.

yaay! thank you very much! I will check out your script today. i was hoping for such a way! thank you pustul!!

So it will probably take some time. That was the version I used too.

He just did less than 2 weeks ago.

I seem to get better performance with DSTM than EWBF. It also lets the card overclock slightly higher (clock +150mhz vs. +135mhz). But the end result on the wallet side seems less than EWBF.

Here are my results with different miners: https://docs.google.com/spreadsheets/d/1-nt-qrBJWz9gXytLmJqtNVG75hjuIXHzbw1eAqvXHEo/edit?usp=sharing

Any plans for a new release soon?

the fee is 2% per day. 72 seconds per hour. all the time

1.
Имя cбoйнoгo пpилoжeния: zm.exe, вepcия: 0.0.0.0, мeткa вpeмeни: 0x5a482db0
Имя cбoйнoгo мoдyля: ntdll.dll, вepcия: 10.0.14393.1715, мeткa вpeмeни: 0x59b0d03e
Кoд иcключeния: 0xc0000005
Cмeщeниe oшибки: 0x000000000002f7fb
Идeнтификaтop cбoйнoгo пpoцecca: 0xc60
Bpeмя зaпycкa cбoйнoгo пpилoжeния: 0x01d3aaded6dc5dbb
Пyть cбoйнoгo пpилoжeния: C:\mining\zm_0.5.8_win\zm.exe
Пyть cбoйнoгo мoдyля: C:\Windows\SYSTEM32\ntdll.dll
Идeнтификaтop oтчeтa: 9b0a6f00-2514-49d8-88b6-43cb24a00e98
Пoлнoe имя cбoйнoгo пaкeтa:
Кoд пpилoжeния, cвязaннoгo co cбoйным пaкeтoм:

2.
Кoнтeйнep oшибки , тип 0
Имя coбытия: APPCRASH
Oтклик: Heт дaнныx
Идeнтификaтop CAB: 0

Cигнaтypa пpoблeмы:
P1: zm.exe
P2: 0.0.0.0
P3: 5a482db0
P4: ntdll.dll
P5: 10.0.14393.1715
P6: 59b0d03e
P7: c0000005
P8: 000000000002f7fb
P9:
P10:

Bлoжeнныe фaйлы:

Эти фaйлы мoжнo нaйти здecь:
C:\ProgramData\Microsoft\Windows\WER\ReportQueue\AppCrash_zm.exe_8412b9179f2a7275c57ff6294b49442c5b66ed25_4df4135b_085951c5

Cимвoл aнaлизa:
Пoвтopный пoиcк peшeния: 0
Идeнтификaтop oтчeтa: 9b0a6f00-2514-49d8-88b6-43cb24a00e98
Cocтoяниe oтчeтa: 100
Xэшиpoвaнный кoнтeйнep:

3.
Имя cбoйнoгo пpилoжeния: zm.exe, вepcия: 0.0.0.0, мeткa вpeмeни: 0x00000000
Имя cбoйнoгo мoдyля: zm.exe, вepcия: 0.0.0.0, мeткa вpeмeни: 0x00000000
Кoд иcключeния: 0xc0000005
Cмeщeниe oшибки: 0x0000000000082381
Идeнтификaтop cбoйнoгo пpoцecca: 0x17c0
Bpeмя зaпycкa cбoйнoгo пpилoжeния: 0x01d3aa54e8933a99
Пyть cбoйнoгo пpилoжeния: C:\mining\zm_0.6_win\zm.exe
Пyть cбoйнoгo мoдyля: C:\mining\zm_0.6_win\zm.exe
Идeнтификaтop oтчeтa: a50aa8d5-6a53-457f-bc2b-00a1dc5cf8a7

4. Last log:

2018-02-25 7:53:24|   GPU5  41C  Sol/s: 467.5  Sol/W: 4.11  Avg: 463.1  I/s: 248.2  Sh: 1.28   0.97 106
2018-02-25 7:53:24|   ========== Sol/s: 2715.3 Sol/W: 3.87  Avg: 2710.5 I/s: 1452.7 Sh: 6.66   0.99 103
2018-02-25 7:53:27|#  connection closed by server r:0
2018-02-25 7:53:27|#  reconnecting
2018-02-25 7:53:37|#  connected to: zec-eu1.nanopool.org:6666
2018-02-25 7:53:39|#  server set difficulty to: 00028f5c0000000000000000...
2018-02-25 14:29:01|#  recv failed: 10054
2018-02-25 14:29:01|#  reconnecting

5. zm.bat
:restart
timeout /T 30
set put=%cd%\
set str=%date%_%time:~0,8%.log
set str=%str::=.%
set str=%str: =0%
set logname=%cd%\%str%
echo %logname%
zm.exe --server zec-eu1.nanopool.org  --port 6666 --user my_user --pass x --telemetry=local_IP:PORT --logfile=%logname%
goto :restart

^
This is very clever! Thanks for the script. @Pustul

The Universe needs you. 

You can use this little python script that I made to handle the miner:


Copy/paste the code in a file, make it executable and run it with "./filename your_launch_command". With a file named "start.py" and for the default dstm's launch script the command would be "./start.py ./zm --cfg-file zm.cfg".

The script will restart the miner when it sees the "cudaMemcpy 1 failed" error message.

BTW, it should work on Windows too if you have python3 installed. Remove the first line and replace "sys.argv[1:]" by "['zm.exe', '--cfg-file', 'zm.cfg']" (or whatever your launch command is).

@dstm it is not software problem, same with simple http request, after about ~24-48 hrs telemetry service stops working, then i simply close and restart cmd (not miner OS) and it works instantly. It is repeating poblem and came with 0.6, was rock solid for month on previous version.

BR.

hi,
just wanted to ask, the fee is 2% per day? or 2% 24 minutes a day?
just for knowledge..

I'm no pro, but I've done a fair bit of digging into protocols, miners, proxies and pool source code. I think that you shouldn't expect to see the exact same adjustment in share submissions when difficulty changes. Think of difficulties as representing submission speed 'brackets'. You'll be set to difficulty X if you submit between n and (n+m) shares - once you exceed (n+m), you'll be set into the next difficulty bracket Y. That sort of thing.


I'm not positive, but it looks like flypool (zcash at least?) isn't vardiff-enabled (variable difficulty)? Meaning, it doesn't dynamically adjust the miner's difficulty, although it allows you to set your own difficulty on miner startup via the password field (defaults to 2000). Can someone correct me if I'm wrong?

Either way, I'm not sure on this. I do know that you shouldn't try to calculate stats on only a minute or two, however. All the advice I've seen is to run a 24+ hour test - even that isn't precise unless you can it on the same hardware at the same time! 


As I mentioned, I'm not sure on flypool being vardiff... I'm interested to hear what you find out, as you dig though!


These are good questions! Keep asking them!

Hello guys,

I am mining in Zcash EWBF but want to change to dstm, but everytime I try the same problem appears: "DSTM exited, waiting to cooldown a bit". How can I connect to the server? anyone experiencing the same problem?

Thanks in advanced,

For me this software runs rocks solid with 1080ti gpus (FTW3 and xtreme water cooled) under Windows 10.

been using dstm since 0.6 on win 10. no crashes. just keep sane clocks and power limits. its actually been more stable than ewbf with the same clocks/power limits.

there is much more to a stable rig than the choice miner software..

Been using it for 3 months straight and NOT ONE single crash. Get off Winblows and use a real OS for server purposes.

That is why you should stay with EWBF, at least for now. Many people complain about DSTM stability. And while it offers 2-8% increase in hashrate (depending on the cards) and a few W per card less power consumption, after subtracting 2% fee and constant crashes (you lose time on restarts) it appears to be less profitable than EWBF (at least for me).

//ZM's vs EWBF//

I know about the difficulty adjustments, but...
1. How the pool knows how good is the miner to find solutions? Isn't the hashrate the indicator? So... if the hashrate increseas with 6%, from 1700 to 1800sol/s, there should be a diff adjustement of 6%, so let's say an accepted share drop from 5/min to 4/min. So how can you explain the accepted shares drop from 6 to 2, between the 2 miners?
2. At start, zcash-flypool begines at 2000 diff. If I got this corectly, it's just the starting difficulty, is not a fix difficulty. So the both miners start at the same diff, so the accepted shares must be greater for ZM with the bigger hashrate at least in the first minute, untill new work is received and the diff readjusted. But guess what... EWBF still beats ZM even at the same diff. I see 4-5 accepteds with EWBF, but only 1 or none with ZM in the first minute.
3.The "more rejecteds" and "more crashes" problems still remains for ZM.
4.With the dev fee on, in ZM (EWBF fee off) difficulty will loose the optimal value with each pool switch (dev fee means ZM switches to dev's pool to mine for him from time to time). So flypool will start readjusting the difficulty after each dev fee, like a new start. How is this more productive?

These are questions from a noob (me). I really want to understand this mining better, not to have online disputes. I hope I'm not missunderstood.
Thanks!

You can't compare share counts/frequency from the miner itself. Depending on a variety of factors (mostly the speed at which you send shares to the pool), the pool will increase the difficulty for your connected miner. Increased difficulty means you find fewer shares, BUT the shares you do find are 'worth more' to the pool (because they were more difficult to find). The overall value of the shares you submit ought to be the same, and the payout the same in those circumstances.

If a miner is better at finding solutions (like dstm is, compared to ewbf), its difficulty will be raised such that it submits fewer shares (lowering the networking requirements) - I'd say this is why when visually comparing ewbf to dstm, you'll see fewer shares being submitted by dstm.

//ZM-0.6 Win vs. EWBF-0.3.4b-BTG-edition Win//

Why the accepted shares are fewer then EWBF's, even though the hashrate is bigger?

My OC settings are not so high, just core +100, power 85, temp 83, mem 0, on 6 GTX 1060 3GB hynix. Core stays under 2000MHz.
My hashrate is around 1800 with DSTM and around 1700 with EWBF. The ping time 63ms, on ZEC-flypool and ZCL-suprnova.
In EWBF I don't get rejected shares aven after 24h, no crashes, and constant acceped shares, around 5 per minute.
In DSTM I see from time to time a rejected, but the main problem is that there are very few accepted shares, like 1-2 per minute.
Why is that? I can't compare the profits because they varry from day to day. I should have 2 rigs with the same setup, but I only have 1 with cards from 3 brands.

DSTM's zm crashes more often than EWBF's with higher OC settings. With EWBF I can go with +100/+500 for a stable mining, but DSTM crashes after a few hours.
My temps are great, 50-52C, the room temp is around 10-12C, the fans are at 50%, so the temp is not the issue. I think ZM can't mannage the higher oc and becomes unstable.
This is strange for a miner that is well payed (in EWBF you can set --fee 0, but not in DSTM) and is 1 year newer than the other, with active developement, so it can benefits from new drivers and new mining techniques.

I realy want to support a miner in active developement, but with smaller performance than other options, I can't. It's all about profits afterall.