Bitcoin Forum
May 23, 2018, 03:54:18 PM *
News: Latest stable version of Bitcoin Core: 0.16.0  [Torrent]. (New!)
 
   Home   Help Search Donate Login Register  
Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 [13] 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 ... 163 »
  Print  
Author Topic: [ANN] dstm's ZCash / Equihash Nvidia Miner v0.6.1 (Linux / Windows)  (Read 205234 times)
dstm
Full Member
***
Offline Offline

Activity: 308
Merit: 119


View Profile
September 30, 2017, 03:46:14 AM
 #241

Thanks you dstm Smiley, now windows version would be really nice Smiley

Right, I've bought it already and installed it on my pc, so the hardest part is done Smiley
1527090858
Hero Member
*
Offline Offline

Posts: 1527090858

View Profile Personal Message (Offline)

Ignore
1527090858
Reply with quote  #2

1527090858
Report to moderator
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction. Advertise here.
1527090858
Hero Member
*
Offline Offline

Posts: 1527090858

View Profile Personal Message (Offline)

Ignore
1527090858
Reply with quote  #2

1527090858
Report to moderator
cryptoDahu
Newbie
*
Offline Offline

Activity: 11
Merit: 0


View Profile WWW
September 30, 2017, 05:47:05 AM
 #242

Right, I've bought it already and installed it on my pc, so the hardest part is done Smiley
nice to hear it!
good work @dstm  Cool
dstm
Full Member
***
Offline Offline

Activity: 308
Merit: 119


View Profile
September 30, 2017, 01:50:45 PM
 #243

I haven't stated it clearly - some users were confused. All telemetry values are averages. Temperature is an exception, it shows the current temp.

The use case for telemetry I had in mind is like this: You don't look at it all the time. You check it let's say 5 times a day. So the averages are a much more accurate overview of the system than the jumpy values of the last run.
jmayniac
Jr. Member
*
Online Online

Activity: 67
Merit: 0


View Profile
October 01, 2017, 02:29:18 AM
 #244

I haven't stated it clearly - some users were confused. All telemetry values are averages. Temperature is an exception, it shows the current temp.

The use case for telemetry I had in mind is like this: You don't look at it all the time. You check it let's say 5 times a day. So the averages are a much more accurate overview of the system than the jumpy values of the last run.

How about some color in the output, it will help to distinguish between items.  Right now with all the text it is hard to see data at a glance.
Nokia555
Sr. Member
****
Offline Offline

Activity: 392
Merit: 266

EthMonitoring.com


View Profile WWW
October 01, 2017, 07:38:00 AM
 #245

Hey dstm,

Is json-rpc missing shares? Accepted, Rejected.

dstm
Full Member
***
Offline Offline

Activity: 308
Merit: 119


View Profile
October 01, 2017, 08:15:21 AM
 #246

Hey dstm,

Is json-rpc missing shares? Accepted, Rejected.


Quote
This is a testing version.... Telemetry is pretty simple currently, if there is anything more you need - suggestions are welcome.

Will add. Smiley
Nokia555
Sr. Member
****
Offline Offline

Activity: 392
Merit: 266

EthMonitoring.com


View Profile WWW
October 01, 2017, 08:45:59 AM
 #247

Hey dstm,

Is json-rpc missing shares? Accepted, Rejected.


Quote
This is a testing version.... Telemetry is pretty simple currently, if there is anything more you need - suggestions are welcome.

Will add. Smiley

If you can, add also fan speed, very useful for monitoring.

dstm
Full Member
***
Offline Offline

Activity: 308
Merit: 119


View Profile
October 01, 2017, 08:50:53 AM
 #248

Hey dstm,

Is json-rpc missing shares? Accepted, Rejected.


Quote
This is a testing version.... Telemetry is pretty simple currently, if there is anything more you need - suggestions are welcome.

Will add. Smiley

If you can, add also fan speed, very useful for monitoring.

Thx Smiley, Nvidia provides this information, will add.
andi_21
Newbie
*
Offline Offline

Activity: 6
Merit: 0


View Profile
October 01, 2017, 10:19:11 AM
 #249

New Version 0.5

con: support set_extranonce rpc
con: improve handling of temporary slow network conditions
con: add monitoring support using web browser
con: add monitoring support using json-rpc
mp: rebalance queue sizes - this improves the solution rate as
seen by the pool, especially on pool that submit new jobs often


This is a testing version, it has a lot internal changes and is less tested. Feedback on stability performance is welcome. Rebalanced queue sizes improve the solution rate on pool side, however this might reduce the performance in situations of heavy cpu load - pls check if there are any improvement on the pool side for you. Telemetry is pretty simple currently, if there is anything more you need - suggestions are welcome.

new version works fine and stable. API works fine as well.
- could you please add current values in JSON as well and not only average values. ( i put all values into grafana for monitoring and statistics)
- lower sols/s on poolside is fixed. localy i have ~2644 and currently (after 24 hours) on flypool up to 2600- (even less than the devfee of 2%). So average Sols/s on poolside is now perfect. very well done.

kryztoval
Newbie
*
Offline Offline

Activity: 12
Merit: 0


View Profile
October 01, 2017, 10:47:35 AM
 #250

I keep getting disconnects every hour or so (as you can see in the graphs) and that happened when I upgraded to 0.4.5

I am now trying 0.5

https://image.prntscr.com/image/4YGPv_YCTyiD0A7l2SfRHg.png
https://image.prntscr.com/image/limqbGRARyyb2riP2Z68RA.png

This is really bad for me since the average with this is a lot lower than it should.
Is there a way to just make just one connection instead of one connection per card?

edit:changed image links
dstm
Full Member
***
Offline Offline

Activity: 308
Merit: 119


View Profile
October 01, 2017, 11:58:05 AM
 #251

I keep getting disconnects every hour or so (as you can see in the graphs) and that happened when I upgraded to 0.4.5

I am now trying 0.5




This is really bad for me since the average with this is a lot lower than it should.
Is there a way to just make just one connection instead of one connection per card?


Thx, if there are any regressions on new versions pls report them as soon as possible, it's welcome and speedups development.

Regular disconnects are bad ofc I'm not sure why you're getting them. It's most likely not an issue on the pool side, since flypool works fine, without disconnects for other users.

All GPUs are separated/independent, this design has a lot of benefits. So for example if you have different GPUs with different speeds in one system - each GPU will receive it's own job and it's own difficulty according to it's speed. Currently I'm not taking full advantage of this design, but it's planned to restarts each GPU separately on connection, hardware etc. failures.

If you have the same issues on 0.5 I'll think about a way to debug this issue, just report it.

Edit: Disconnects happen very regular, after the same amount of time, that's very strange.
dstm
Full Member
***
Offline Offline

Activity: 308
Merit: 119


View Profile
October 01, 2017, 12:17:55 PM
 #252

New Version 0.5

con: support set_extranonce rpc
con: improve handling of temporary slow network conditions
con: add monitoring support using web browser
con: add monitoring support using json-rpc
mp: rebalance queue sizes - this improves the solution rate as
seen by the pool, especially on pool that submit new jobs often


This is a testing version, it has a lot internal changes and is less tested. Feedback on stability performance is welcome. Rebalanced queue sizes improve the solution rate on pool side, however this might reduce the performance in situations of heavy cpu load - pls check if there are any improvement on the pool side for you. Telemetry is pretty simple currently, if there is anything more you need - suggestions are welcome.

new version works fine and stable. API works fine as well.
- could you please add current values in JSON as well and not only average values. ( i put all values into grafana for monitoring and statistics)
- lower sols/s on poolside is fixed. localy i have ~2644 and currently (after 24 hours) on flypool up to 2600- (even less than the devfee of 2%). So average Sols/s on poolside is now perfect. very well done.




Quote
                                                                                                                       
could you please add current values in JSON as well and not only average values. ( i put all values into grafana for monitoring and statistics)
                                                                                                                       
Ofc, will add them.
I was also thinking about adding plots of current values to the web output, however this seemed useless for me since they are noisy / jumping randomly, average values are the things that really matter. So I kept it simple which has advantages if you display the things on mobile devices.

Quote
                                                                                                                       
lower sols/s on poolside is fixed. localy i have ~2644 and currently (after 24 hours) on flypool up to 2600- (even less than the devfee of 2%). So average Sols/s on poolside is now perfect. very well done.                                               
                                                                                                                       
Nice, as I said, my numbers are exact even slightly rounded down Smiley
kryztoval
Newbie
*
Offline Offline

Activity: 12
Merit: 0


View Profile
October 02, 2017, 03:51:34 AM
 #253

Thx, if there are any regressions on new versions pls report them as soon as possible, it's welcome and speedups development.

Regular disconnects are bad ofc I'm not sure why you're getting them. It's most likely not an issue on the pool side, since flypool works fine, without disconnects for other users.

All GPUs are separated/independent, this design has a lot of benefits. So for example if you have different GPUs with different speeds in one system - each GPU will receive it's own job and it's own difficulty according to it's speed. Currently I'm not taking full advantage of this design, but it's planned to restarts each GPU separately on connection, hardware etc. failures.

If you have the same issues on 0.5 I'll think about a way to debug this issue, just report it.

Edit: Disconnects happen very regular, after the same amount of time, that's very strange.

I switched miners to ewbf to test, and the graph I got was no more disconnections. This has to be some sort of throttling, I have 24 cards, I split them in groups of equals. So tot he pool it must seem like I am making 24 separate connections at the exact same time.

you can see the difference between zm disconnecting, and ewbf staying normalized. Something is amiss, but I don't know exactly what it is, nothing changed, only the miner program.
https://image.prntscr.com/image/SkRIFTAmQOWRi5WTz_2ang.png
dstm
Full Member
***
Offline Offline

Activity: 308
Merit: 119


View Profile
October 02, 2017, 10:48:52 AM
 #254

Thx, if there are any regressions on new versions pls report them as soon as possible, it's welcome and speedups development.

Regular disconnects are bad ofc I'm not sure why you're getting them. It's most likely not an issue on the pool side, since flypool works fine, without disconnects for other users.

All GPUs are separated/independent, this design has a lot of benefits. So for example if you have different GPUs with different speeds in one system - each GPU will receive it's own job and it's own difficulty according to it's speed. Currently I'm not taking full advantage of this design, but it's planned to restarts each GPU separately on connection, hardware etc. failures.

If you have the same issues on 0.5 I'll think about a way to debug this issue, just report it.

Edit: Disconnects happen very regular, after the same amount of time, that's very strange.

I switched miners to ewbf to test, and the graph I got was no more disconnections. This has to be some sort of throttling, I have 24 cards, I split them in groups of equals. So tot he pool it must seem like I am making 24 separate connections at the exact same time.

you can see the difference between zm disconnecting, and ewbf staying normalized. Something is amiss, but I don't know exactly what it is, nothing changed, only the miner program.



What zm version is this?

Disconnects are very regular. Is there anything that happens every hour on your host/network? It's very strange that it happens exactly after one hour but works fine 1 hour long. Other users seem not to have this issues.                                 
24 simultaneous connection is a nothing special, browsers open up to 8/16 sim. connections per website/domain.                 
I'll check ofc if there is something wrong in my code.
Could you pls. pm me some logfiles?

Btw. zm looks faster on pool side Smiley
kryztoval
Newbie
*
Offline Offline

Activity: 12
Merit: 0


View Profile
October 02, 2017, 08:07:25 PM
 #255

Quote
What zm version is this?
4.5 and 5

Quote
Disconnects are very regular.
Yes they appear regular, however disconnects only happen when zm is running. Otherwise the computers do not disconnect. I tested this by only running zm in one GPU miner while the others where doing other miners and only the box with zm disconnected.

Quote
Is there anything that happens every hour on your host/network?
Not that I am aware of.

Quote
It's very strange that it happens exactly after one hour but works fine 1 hour long.
It's approximately one hour and it starts counting at the time zm starts running. I don't see any memory leaks. It seems a lot like a build up of connections not being cleaned until network crashes.

Quote
Other users seem not to have this issues.                                 
This is puzzling, sure, I do have 3 boxes with the same behavior. I'll go ahead and see if I can reproduce this behavior in a freshly installed box.

Quote
24 simultaneous connection is a nothing special, browsers open up to 8/16 sim. connections per website/domain.
I agree, but all the connections are closed. And most connections are to servers that are meant for multiple connections (as httpd).

Quote
I'll check ofc if there is something wrong in my code.
Hard to check if we can't pinpoint where to look.

Quote
Could you pls. pm me some logfiles?
Sure, what logs do you need?

Quote
Btw. zm looks faster on pool side Smiley
It most definitively is.
dstm
Full Member
***
Offline Offline

Activity: 308
Merit: 119


View Profile
October 02, 2017, 09:16:30 PM
 #256

Quote
I tested this by only running zm in one GPU miner while the others where doing other miners and only the box with zm disconnected.
                                                                                                                     
We'll get it Smiley The basics first, I hope it costs not too much time Smiley
Are your systems on wifi? If yes, you could try using a cable if possible.
You might test running zm on less GPUs using the '--dev' option so you'll have less connections per system.

Quote
It seems a lot like a build up of connections not being cleaned until network crashes.
Not sure what you mean by 'build up of connections'. It opens one socket per GPU there should be no 'build up of connections'.

Quote
I agree, but all the connections are closed. And most connections are to servers that are meant for multiple connections (as httpd).
I have a 10GPU system running stable, no disconnects, on flypool. I'm pretty sure they have no issues handling much more connections. If we run out of ideas I'll write a test for this Smiley

Quote
Sure, what logs do you need?
Complete output till it disconnects (pls include also the 'connection closed by server' message - if you get it).


Edit:
I'm pretty sure there should be no issues with the number of connections on the pool side. So let's say you have 1000 rigs mining the same ZEC address, the pool must be able to handle all 1000 connections mining the same ZEC address.
miningfarmcad
Newbie
*
Offline Offline

Activity: 1
Merit: 0


View Profile
October 02, 2017, 10:05:33 PM
 #257

Thank you DSTM for your software.

Las week I used it with 2500 GPUs, but I had a lot of problems. 10% of the Gpus crash each 30-60 minutes. I really want to work with you to solve that problem. We plan to switch more GPUs but we need more stability and after we will have specials requests for the JSON API.

Questions :

1- What is the ideal nvidia driver version ? We use the latest : 384.90
2- What is the perfect Ubuntu version ? We use Ubuntu 14.04
3- What log I need to follow to try to find the problem ?


When a GPU fail, could you recover it ? I note with nvidia-smi that the GPU still working. I think EWBF recover failed GPU.
dstm
Full Member
***
Offline Offline

Activity: 308
Merit: 119


View Profile
October 02, 2017, 11:04:55 PM
 #258

Thank you DSTM for your software.

Las week I used it with 2500 GPUs, but I had a lot of problems. 10% of the Gpus crash each 30-60 minutes. I really want to work with you to solve that problem. We plan to switch more GPUs but we need more stability and after we will have specials requests for the JSON API.

Questions :

1- What is the ideal nvidia driver version ? We use the latest : 384.90
2- What is the perfect Ubuntu version ? We use Ubuntu 14.04
3- What log I need to follow to try to find the problem ?


When a GPU fail, could you recover it ? I note with nvidia-smi that the GPU still working. I think EWBF recover failed GPU.


Quote
Las week I used it with 2500 GPUs, but I had a lot of problems. 10% of the Gpus crash each 30-60 minutes. I really want to work with you to solve that problem. We plan to switch more GPUs but we need more stability and after we will have specials requests for the JSON API.
I had no reports about crashes, the development is pretty fast paced currently so there could be bugs ofc. I think the fastest/easiest way to resolve the issues is to have ssh access to one of your systems that crashes.

JSON API: this is pretty easy to extend, currently I've only a basic set, It was meant for testing.

Quote
1- What is the ideal nvidia driver version ? We use the latest : 384.90
2- What is the perfect Ubuntu version ? We use Ubuntu 14.04
3- What log I need to follow to try to find the problem ?

1. I've tested zm on 375.66 and 384.90, both perform equal without issues.
2. I've tested zm on 16.04 so I can't make a robust statement about 14.04
3. It's much faster/easier if you could provide an ssh access to one of your systems.

Quote
When a GPU fail, could you recover it ? I note with nvidia-smi that the GPU still working. I think EWBF recover failed GPU.
ZM is designed such that every GPU is separated and independent, so yes on some cases a crashed GPU is recoverable (not always!), this is currently not implemented but it's planned.
kryztoval
Newbie
*
Offline Offline

Activity: 12
Merit: 0


View Profile
October 03, 2017, 10:07:10 AM
 #259

Quote
Are your systems on wifi? If yes, you could try using a cable if possible.
You might test running zm on less GPUs using the '--dev' option so you'll have less connections per system.
                                                                                                                     
Systems are wired.

Quote
Not sure what you mean by 'build up of connections'. It opens one socket per GPU there should be no 'build up of connections'.
Imagine a memory leak, but made with open sockets. That is what this feels like. could be or not, I am just trying to explain it somehow. Like, if you have a disconect, do you close the socket or does garbage collection close it? do you reuse the sockets? what if the sockets are left alive and for some reason my installation doesn't do garbage collection until network crashes? again, this is pure guesswork. this is a shot in the dark and may not even be possible. Just saying what it "feels" like.

Quote
I have a 10GPU system running stable, no disconnects, on flypool. I'm pretty sure they have no issues handling much more connections. If we run out of ideas I'll write a test for this Smiley
Seems weird. I imaged the system as ubuntu 17.04. I can get the image online if you want to try it or I can get you access to one of the rigs too if this proves too troublesome.

Quote
Complete output till it disconnects (pls include also the 'connection closed by server' message - if you get it).
I'll make it with -dev, give me a couple of hours to let it run and I'll send you those.

Quote
Edit:
I'm pretty sure there should be no issues with the number of connections on the pool side. So let's say you have 1000 rigs mining the same ZEC address, the pool must be able to handle all 1000 connections mining the same ZEC address.
Yeah, I am pretty sure this is something that should not happen, this is weird.
NameTaken
Hero Member
*****
Offline Offline

Activity: 630
Merit: 502


View Profile
October 03, 2017, 12:45:56 PM
 #260

I think an intensity option would be useful for people who want their desktop to remain responsive while mining for example.
Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 [13] 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 ... 163 »
  Print  
 
Jump to:  

Sponsored by , a Bitcoin-accepting VPN.
Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!