3.3M share round is crazy for this difficulty
20M round would be crazy, 10M is not :) 
Topic: [1200 TH] EMC: 0 Fee DGM. Anonymous PPS. US & EU servers. No Registration! - page 164. (Read 499597 times)
December 23, 2011, 01:58:27 PM
December 23, 2011, 01:29:41 PM
3.3M share round is crazy for this difficulty
Active Miners 496 Round Duration 16:17:32
Current Speed 244.65 GH/s Round Shares 3335104
OK, sunbreak, your do for a block...
December 23, 2011, 10:00:12 AM
Omo - I have made the changes you suggested.
I saw it, thank you!
December 23, 2011, 03:03:39 AM
You might ask the mod if moving these last bunch of message to the cgminer thread is possible - it would make sense and be more considerate here.
I just more-or-less duplicated my previous post over there.
December 23, 2011, 02:53:31 AM
You might ask the mod if moving these last bunch of message to the cgminer thread is possible - it would make sense and be more considerate here.
December 23, 2011, 02:49:52 AM
Ah but you compile them again yourself right?
I do both - compile myself - and also run the binary that ckolivas releases.
Thus I know if there are any issues ... which is what lead to the extra comment at the end of my script about linking the missing library.
It failed for me (on 11.04) so I worked out what was needed to fix that.
Edit: and I have finally got around to creating a Virtual 11.10 to compile with so I can do that myself also (I was compiling on 11.04 on an older PC)
So I guess I'll also create a new version of my script soon when I bother to create a new 11.10 USB (since 11.10 has been around for a while ...)
I do both - compile myself - and also run the binary that ckolivas releases.
Thus I know if there are any issues ... which is what lead to the extra comment at the end of my script about linking the missing library.
It failed for me (on 11.04) so I worked out what was needed to fix that.
Edit: and I have finally got around to creating a Virtual 11.10 to compile with so I can do that myself also (I was compiling on 11.04 on an older PC)
So I guess I'll also create a new version of my script soon when I bother to create a new 11.10 USB (since 11.10 has been around for a while ...)
I did both with 2.0.8, but compiled myself for the large majority of the experiments. With 2.0.7 I used the distributed binary.
December 23, 2011, 02:48:12 AM
... The only time I've ever had lock ups is when gpu settings were pushed too far. Engine 950 is not that high but some cards will choke on even that. I'd definitely start really conservative at 850 and default mem speed and then see how stability is before altering any gpu settings.
Cards behave very differently even from the same mfr. batch. I have 2 5830s bought together and one craps out above 970 while the other is happy at 1020. If you got a fussy card then anything higher than 850 could be causing failures. If cgminer starts and outputs init messages then I'd say software is likely ok and you should debug hardware issues and really back off to ultra conservative values, and no "auto" settings that may be causing cgminer to adjust things. Also start with one card and add cards as stability is verified.
Cards behave very differently even from the same mfr. batch. I have 2 5830s bought together and one craps out above 970 while the other is happy at 1020. If you got a fussy card then anything higher than 850 could be causing failures. If cgminer starts and outputs init messages then I'd say software is likely ok and you should debug hardware issues and really back off to ultra conservative values, and no "auto" settings that may be causing cgminer to adjust things. Also start with one card and add cards as stability is verified.
I changed gpu-engine from "0-950" to "850" and removed both "auto" settings. That caused it to go a few more seconds -- to the extent of outputing three "Accepted" messages that it hadn't previously...
[2011-12-22 22:38:28] Long-polling activated for http://us.eclipsemc.com:8337/LP
[2011-12-22 22:38:32] Accepted 00000000.943f188b.123aad4f GPU 2 thread 2 pool 0
GPU2 47.5C 4504RPM | (5s):326.7 (avg):4026.5 Mh/s | A:1 R:0 HW:0 U:600.00/m I:8
[2011-12-22 22:38:33] Accepted 00000000.ddd1a27e.08200eba GPU 1 thread 1 pool 0
GPU1 52.5C | (5s):363.6 (avg):222.7 Mh/s | A:1 R:0 HW:0 U:11.99/m I:5
[2011-12-22 22:38:34] Accepted 00000000.307282d1.c779af85 GPU 2 thread 2 pool 0
GPU2 48.0C 4519RPM | (5s):333.0 (avg):238.0 Mh/s | A:2 R:0 HW:0 U:23.98/m I:8
...but then the system was frozen.
So I think we might be on a productive track. Should we take this to another forum, like the cgminer forum? I don't want to slobber all over the Eclipse topic with my pool-independent issues...
December 23, 2011, 02:34:11 AM
Ah but you compile them again yourself right?
I do both - compile myself - and also run the binary that ckolivas releases.
Thus I know if there are any issues ... which is what lead to the extra comment at the end of my script about linking the missing library.
It failed for me so I worked out what was needed to fix that.
Edit: and I have finally got around to creating a Virtual 11.10 to compile with so I can do that myself also (I was compiling on 11.04 on an older PC)
So I guess I'll also create a new version of my script soon when I bother to create a new 11.10 USB (since 11.10 has been around for a while ...)
I'm still using 11.04 but updating that, not upgrading to the newer release. Not sure if that's why but haven't seen anything weird. I didn't recompile cgminer until last night when I merged my code changes from 2.0.6 to 2.0.8.I do both - compile myself - and also run the binary that ckolivas releases.
Thus I know if there are any issues ... which is what lead to the extra comment at the end of my script about linking the missing library.
It failed for me so I worked out what was needed to fix that.
Edit: and I have finally got around to creating a Virtual 11.10 to compile with so I can do that myself also (I was compiling on 11.04 on an older PC)
So I guess I'll also create a new version of my script soon when I bother to create a new 11.10 USB (since 11.10 has been around for a while ...)
(I didn't update any of my systems to 11.10 as an act of rebellion against the choice to force us into Unity by default. I know I can remove it and switch back to the "classic" desktop. I just don't like this MS-like attitude of forcing all sorts of changes on us "classic" users. It irritated me when they started putting the [X] button on the left side of windows. /rant )
December 23, 2011, 02:25:03 AM
Ah but you compile them again yourself right?
I do both - compile myself - and also run the binary that ckolivas releases.
Thus I know if there are any issues ... which is what lead to the extra comment at the end of my script about linking the missing library.
It failed for me (on 11.04) so I worked out what was needed to fix that.
Edit: and I have finally got around to creating a Virtual 11.10 to compile with so I can do that myself also (I was compiling on 11.04 on an older PC)
So I guess I'll also create a new version of my script soon when I bother to create a new 11.10 USB (since 11.10 has been around for a while ...)
I do both - compile myself - and also run the binary that ckolivas releases.
Thus I know if there are any issues ... which is what lead to the extra comment at the end of my script about linking the missing library.
It failed for me (on 11.04) so I worked out what was needed to fix that.
Edit: and I have finally got around to creating a Virtual 11.10 to compile with so I can do that myself also (I was compiling on 11.04 on an older PC)
So I guess I'll also create a new version of my script soon when I bother to create a new 11.10 USB (since 11.10 has been around for a while ...)
December 23, 2011, 02:19:30 AM
You might not have seen my subsequent post before you replied, but anyway, I tried 11.04 and 2.4 last night. I guess I violated the advice "don't update anything" because I did update immediately after installing.
I don't think that updating should have much effect as I keep my 11.04 right up to date and haven't seen any problems. The only time I've ever had lock ups is when gpu settings were pushed too far. Engine 950 is not that high but some cards will choke on even that. I'd definitely start really conservative at 850 and default mem speed and then see how stability is before altering any gpu settings.Cards behave very differently even from the same mfr. batch. I have 2 5830s bought together and one craps out above 970 while the other is happy at 1020. If you got a fussy card then anything higher than 850 could be causing failures. If cgminer starts and outputs init messages then I'd say software is likely ok and you should debug hardware issues and really back off to ultra conservative values, and no "auto" settings that may be causing cgminer to adjust things. Also start with one card and add cards as stability is verified.
December 23, 2011, 02:12:28 AM
So ... after ignoring my suggestions for so long ... why not do what I said to start with 
Install 11.04 and 2.4 - (as per my sig) - what I've been using since forever (July) and still running fine
... and again ... don't update anything.
It's worked with every version of cgminer since I started - 1.3.1 all the way up to 2.0.8
The only change required since then is also in my script at the end: the latest releases are compiled in 11.10 and there is a name change in one library in 11.10.
Install 11.04 and 2.4 - (as per my sig) - what I've been using since forever (July) and still running fine
... and again ... don't update anything.
It's worked with every version of cgminer since I started - 1.3.1 all the way up to 2.0.8
The only change required since then is also in my script at the end: the latest releases are compiled in 11.10 and there is a name change in one library in 11.10.
You might not have seen my subsequent post before you replied, but anyway, I tried 11.04 and 2.4 last night. I guess I violated the advice "don't update anything" because I did update immediately after installing.
December 23, 2011, 01:57:16 AM
In many days of trying, with cgminer 2.0.7 and 2.0.8, with Ubuntu 10.04, 11.04, and 11.10 and Catalysts ranging from 11.6 to 11.12 and AMD SDKs from 2.4 to 2.6 I have never gotten cgminer to go more than a few seconds without hanging (best case) or freezing the entire system (worst case; so far, no smoke).
So ... after ignoring my suggestions for so long ... why not do what I said to start with Install 11.04 and 2.4 - (as per my sig) - what I've been using since forever (July) and still running fine
... and again ... don't update anything.
It's worked with every version of cgminer since I started - 1.3.1 all the way up to 2.0.8
The only change required since then is also in my script at the end: the latest releases are compiled in 11.10 and there is a name change in one library in 11.10.
December 23, 2011, 01:47:21 AM
Ha welcome to bitcoin mining 
. I spent 2 months tweaking and testing software/hardware before any real success. Innaba has a couple great linux guides, and Kano has a good guide on cgminer install but the quickest easiest solution I found is BAMT https://bitcointalksearch.org/topic/bamt-easy-persistent-usb-key-based-linux-for-dedicated-minersmining-farms-28967. As far as cgminer, the most stable setup I've found with my 5970 rig is 11.04 with catalyst 11.6 and cgminer 2.0.7 (using the 64-built.tar.bz2). Everything else I've tried has not worked well. I ended up wiping 2 rigs and doing a fresh install with no problems since.
. I spent 2 months tweaking and testing software/hardware before any real success. Innaba has a couple great linux guides, and Kano has a good guide on cgminer install but the quickest easiest solution I found is BAMT https://bitcointalksearch.org/topic/bamt-easy-persistent-usb-key-based-linux-for-dedicated-minersmining-farms-28967. As far as cgminer, the most stable setup I've found with my 5970 rig is 11.04 with catalyst 11.6 and cgminer 2.0.7 (using the 64-built.tar.bz2). Everything else I've tried has not worked well. I ended up wiping 2 rigs and doing a fresh install with no problems since.I've seen and used the guides. I just tried 11.04 with Catalyst 11.6 and the pre-built cgminer 2.0.7, i.e., followed your suggestion, and had the same problem I've had with every other combo.
But thanks! (no sarcasm, I do appreciate any attempt to help) What I've found from all the good-intentioned advice to "run what I'm running" is that a lot of people run cgminer successfully, which was obvious anyway, and that I have some weird problem likely caused by a stupid error or omission on my part that prevents my running it. So, no more (re)installing Ubuntu/Catalyst/SDK/cgminer versions; unless someone comes up with a specific thing to try based on a plausible rationale, I'm moving on to other mining software. Since my technical curiosity is piqued, I'm putting investigating my problem with cgminer using GDB on my list of things to do, but it's not a high priority right now.
December 23, 2011, 01:36:44 AM
I would go with cgminer, it clocks and handles the fans all in one, while monitoring the temps.
The way I got cgminer to work was create a json config file with my params and that is the only argument I call cgminer with
./cgminer -c miner.emc.json
I've been starting it from a ssh session using a script that contains:
Code:
xhost + &> /dev/null
now="`date +%Y.%m.%d.%H.%M.%S`"
cd ~/miners/cgminer
DISPLAY=:0 cgminer -c cgminer.conf 2> logs/$now.log
now="`date +%Y.%m.%d.%H.%M.%S`"
cd ~/miners/cgminer
DISPLAY=:0 cgminer -c cgminer.conf 2> logs/$now.log
The cgminer that I'm running is in a /bin that's on my PATH. I don't see any significant difference from what you're doing.
Quote
I have 4 rigs with 4X5970's. I run them at 820/420. (sometimes 800/420) intensity 7. here is the config file:
Code:
{
"_comment1" : "Any long-format command line argument ",
"_comment2" : "may be used in this JSON configuration file",
"url" : "http://us.eclipsemc.com:8337",
"user" : "username.miner4",
"pass" : "password",
"gpu-engine": "820",
"gpu-memclock": "420",
"auto-fan": true,
"intensity": "7"
}
"_comment1" : "Any long-format command line argument ",
"_comment2" : "may be used in this JSON configuration file",
"url" : "http://us.eclipsemc.com:8337",
"user" : "username.miner4",
"pass" : "password",
"gpu-engine": "820",
"gpu-memclock": "420",
"auto-fan": true,
"intensity": "7"
}
Here's mine:
Code:
{
"pools" : [
{
"url" : "http://us.eclipsemc.com:8337",
"user" : "username-here",
"pass" : "password-here"
},
{
"url" : "http://192.168.168.103:8332",
"user" : "brec",
"pass" : "password-here"
}
],
"intensity" : "d,d,8,8,8,8",
"gpu-engine" : "0-950",
"gpu-fan" : "0-85",
"temp-cutoff" : "95,95,95,95,95,95",
"temp-overheat" : "85,85,85,85,85,85",
"temp-target" : "75,75,75,75,75,75",
"auto-fan" : true,
"auto-gpu" : true,
"expiry" : "120",
"gpu-threads" : "2",
"log" : "5",
"queue" : "1",
"retry-pause" : "5",
"scan-time" : "60",
"temp-hysteresis" : "3",
"worksize" : "0",
"donation" : "0.00",
"shares" : "0",
"kernel-path" : "/home/brec/miners/cgminer"
}
"pools" : [
{
"url" : "http://us.eclipsemc.com:8337",
"user" : "username-here",
"pass" : "password-here"
},
{
"url" : "http://192.168.168.103:8332",
"user" : "brec",
"pass" : "password-here"
}
],
"intensity" : "d,d,8,8,8,8",
"gpu-engine" : "0-950",
"gpu-fan" : "0-85",
"temp-cutoff" : "95,95,95,95,95,95",
"temp-overheat" : "85,85,85,85,85,85",
"temp-target" : "75,75,75,75,75,75",
"auto-fan" : true,
"auto-gpu" : true,
"expiry" : "120",
"gpu-threads" : "2",
"log" : "5",
"queue" : "1",
"retry-pause" : "5",
"scan-time" : "60",
"temp-hysteresis" : "3",
"worksize" : "0",
"donation" : "0.00",
"shares" : "0",
"kernel-path" : "/home/brec/miners/cgminer"
}
The second "pool" is bitcoind on my Mac.
In many days of trying, with cgminer 2.0.7 and 2.0.8, with Ubuntu 10.04, 11.04, and 11.10 and Catalysts ranging from 11.6 to 11.12 and AMD SDKs from 2.4 to 2.6 I have never gotten cgminer to go more than a few seconds without hanging (best case) or freezing the entire system (worst case; so far, no smoke).
December 22, 2011, 08:40:27 PM
Omo - I have made the changes you suggested.
December 22, 2011, 07:06:10 PM
Subject: Newbie seeks guidance
In a nutshell: I decided to mine, spent many hours reading bitcointalk.org and elsewhere, built a 3x5970 rig, decided on cgminer, decided on Eclipse. I've spent days trying to get cgminer to work on Ubuntu without success...
In a nutshell: I decided to mine, spent many hours reading bitcointalk.org and elsewhere, built a 3x5970 rig, decided on cgminer, decided on Eclipse. I've spent days trying to get cgminer to work on Ubuntu without success...
Ha welcome to bitcoin mining
. I spent 2 months tweaking and testing software/hardware before any real success. Innaba has a couple great linux guides, and Kano has a good guide on cgminer install but the quickest easiest solution I found is BAMT https://bitcointalksearch.org/topic/bamt-easy-persistent-usb-key-based-linux-for-dedicated-minersmining-farms-28967. As far as cgminer, the most stable setup I've found with my 5970 rig is 11.04 with catalyst 11.6 and cgminer 2.0.7 (using the 64-built.tar.bz2). Everything else I've tried has not worked well. I ended up wiping 2 rigs and doing a fresh install with no problems since. December 22, 2011, 06:55:57 PM
I would go with cgminer, it clocks and handles the fans all in one, while monitoring the temps.
The way I got cgminer to work was create a json config file with my params and that is the only argument I call cgminer with
./cgminer -c miner.emc.json
I have 4 rigs with 4X5970's. I run them at 820/420. (sometimes 800/420) intensity 7. here is the config file:
Code:
{
"_comment1" : "Any long-format command line argument ",
"_comment2" : "may be used in this JSON configuration file",
"url" : "http://us.eclipsemc.com:8337",
"user" : "username.miner4",
"pass" : "password",
"gpu-engine": "820",
"gpu-memclock": "420",
"auto-fan": true,
"intensity": "7"
}
"_comment1" : "Any long-format command line argument ",
"_comment2" : "may be used in this JSON configuration file",
"url" : "http://us.eclipsemc.com:8337",
"user" : "username.miner4",
"pass" : "password",
"gpu-engine": "820",
"gpu-memclock": "420",
"auto-fan": true,
"intensity": "7"
}
edit: guess I forgot to mention i use linuxcoin on a thumb drive. with a clean thumb drive you can download miner, download a json file, and start the miner with minutes, already has all the drivers loaded.
December 22, 2011, 06:26:54 PM
Subject: Newbie seeks guidance
In a nutshell: I decided to mine, spent many hours reading bitcointalk.org and elsewhere, built a 3x5970 rig, decided on cgminer, decided on Eclipse. I've spent days trying to get cgminer to work on Ubuntu without success (that story is in the cgminer thread starting here), but based on preliminary trials (and some help from Inaba - thanks!) poclbm seems to work. So I have vetted my hardware, or at least two GPU cores thereof, and I think either poclbm or Phoenix will be OK.
Now how do I get from here to a state where all six cores are running more-or-less without intervention? Do I accept that it will be necessary to spend quite a while tuning clock and fan speeds while monitoring temperatures, and that this process will be fairly arduous? That's the way it looks, but maybe that's because I'm ignorant about available tools. On a related note, I'm considering switching from Ubuntu to Windows 7.
Any (good) advice welcome!
In a nutshell: I decided to mine, spent many hours reading bitcointalk.org and elsewhere, built a 3x5970 rig, decided on cgminer, decided on Eclipse. I've spent days trying to get cgminer to work on Ubuntu without success (that story is in the cgminer thread starting here), but based on preliminary trials (and some help from Inaba - thanks!) poclbm seems to work. So I have vetted my hardware, or at least two GPU cores thereof, and I think either poclbm or Phoenix will be OK.
Now how do I get from here to a state where all six cores are running more-or-less without intervention? Do I accept that it will be necessary to spend quite a while tuning clock and fan speeds while monitoring temperatures, and that this process will be fairly arduous? That's the way it looks, but maybe that's because I'm ignorant about available tools. On a related note, I'm considering switching from Ubuntu to Windows 7.
Any (good) advice welcome!
December 22, 2011, 02:25:08 PM
I seem to be getting MANY get work failures or 'Pool 0 not providing work fast enough' as cgminer puts it. In fact 442 in about 8 hours 
Connected to http://us.eclipsemc.com:9009, but also has problems with the default port.
What's up with that?
Other info: located in California. Had no problems with bitcoins.lc pool. Running at a whopping 150MH/s.
Connected to http://us.eclipsemc.com:9009, but also has problems with the default port.
What's up with that?
Other info: located in California. Had no problems with bitcoins.lc pool. Running at a whopping 150MH/s.
Try changing the extra queue items to 5 or more..helped me with the same error. You may also have to lower the scan time if this causes rejects.
December 22, 2011, 02:01:16 PM
I see, thanks 
Jump to: