OFFICIAL CGMINER mining software thread for linux/win/osx/mips/arm/r-pi 4.11.1

[Xian01]

I'm getting a lot of;

KLNx:0 went idle before work was sent

... spam running the 3.6.6 nogpu test executable for Windows. Mining at ghash.io FWIW.

 

[jmc1517]

Edit: Since the restart, the "to" candidate has run for seven hours with no errors or anomalies, still going strong.

Here's some more, trying to narrow it down further then please.

http://ck.kolivas.org/apps/cgminer/temp/cgminer-tob.exe
http://ck.kolivas.org/apps/cgminer/temp/cgminer-tor.exe
http://ck.kolivas.org/apps/cgminer/temp/cgminer-tow.exe

As an aside: I've been running 3.5.1 since yesterday midday. No errors, so I'm still happy with it.

So, I've trialled the new 3 above.  Got zombies fairly quickly on all of them so terminated the tests soon after they occurred (I can leave them running longer if it would help prove anything).

Results as follows:
cgminer-tob started at 09:33, first zombie at 09:37, 2nd at 9:43. Stopped at 09:44.
cgminer-tor  started at 09:48, first zombie at 10:00, 2nd at 10:23. Stopped at 10:41.
cgminer-tow started at 10:45, first zombie AMU 32 after 2 minutes, but must have failed and auto-replugged (?)
                   as AMU 32 is now showing after AMU 33 in the table.

Logfiles here (without --debug):
 https://dl.dropboxusercontent.com/u/44240170/logfile-tob.txt
 https://dl.dropboxusercontent.com/u/44240170/logfile-tor.txt
 https://dl.dropboxusercontent.com/u/44240170/logfile-tow.txt

Currently running cgminer-tow with --debug. Just on the off-chance of capturing an error...
 

[aigeezer]

Edit: Since the restart, the "to" candidate has run for seven hours with no errors or anomalies, still going strong.

Here's some more, trying to narrow it down further then please.

http://ck.kolivas.org/apps/cgminer/temp/cgminer-tob.exe
http://ck.kolivas.org/apps/cgminer/temp/cgminer-tor.exe
http://ck.kolivas.org/apps/cgminer/temp/cgminer-tow.exe

As an aside: I've been running 3.5.1 since yesterday midday. No errors, so I'm still happy with it.

So, I've trialled the new 3 above.  Got zombies fairly quickly on all of them so terminated the tests soon after they occurred (I can leave them running longer if it would help prove anything).

Results as follows:
cgminer-tob started at 09:33, first zombie at 09:37, 2nd at 9:43. Stopped at 09:44.
cgminer-tor  started at 09:48, first zombie at 10:00, 2nd at 10:23. Stopped at 10:41.
cgminer-tow started at 10:45, first zombie AMU 32 after 2 minutes, but must have failed and auto-replugged (?)
                   as AMU 32 is now showing after AMU 33 in the table.

Logfiles here (without --debug):
 https://dl.dropboxusercontent.com/u/44240170/logfile-tob.txt
 https://dl.dropboxusercontent.com/u/44240170/logfile-tor.txt
 https://dl.dropboxusercontent.com/u/44240170/logfile-tow.txt

Currently running cgminer-tow with --debug. Just on the off-chance of capturing an error...
 

Just starting to catch up to you. "tob" crashed within seconds. That unhelpful Windows message "cgminer-nogpu has stopped working". On to "tor".

Edit: AMU7 has gone zombie about 15 minutes in to "tor". I'll try "tow" now.

Edit: "cgminer-nogpu.exe has stopped working" after about 15 seconds in "tow". Back to "tob" for a while, I guess.

Edit: "tob" has a zombie on AMU6 for this run about 20 minutes in. Last time it crashed early. Tough stuff to track down.

Edit: just noticed (2 hours in to "tob" run) that there is no LED on, even though AMU6 is still reported as zombie. That's a first, I think. Sorry - unit wasn't into socket far enough.

Weird - As I'm typing, all but one (see above) of the LEDs came on solid and the BAL fan has slowed. The display is still updating normally, it seems and everything claims to be hashing. I don't trust it though, so I'll restart it.

Edit: Several more zombies from time to time. Another interesting pattern, unique to "tob" perhaps - I usually get the "cgminer-nogpu has stopped working" crash now when I try to restart a zombie by disconnecting either just it  or its whole hub and replugging.

[ksenter]

May have been a coincidence but yesterday I finally updated from 3.6.4 to 3.6.6.  A few hours later I found my PC was locked up, had to hold down the power button.  I switched back to 3.6.4 just in case because I'm going to be away today.  

Running windows 8 x64 with a BFL little single.

[P_Shep]

Do the pool stats have a 'ping' like time to them? Or are my stats not working... I'm getting this all the time:

                    [Calls] => 0
                    [Wait] => 0
                    [Max] => 0
                    [Min] => 99999999
                    [Pool Calls] => 0
                    [Pool Attempts] => 0
                    [Pool Wait] => 0
                    [Pool Max] => 0
                    [Pool Min] => 99999999
                    [Pool Av] => 0

[-ck]

Do the pool stats have a 'ping' like time to them? Or are my stats not working... I'm getting this all the time:

                    [Calls] => 0
                    [Wait] => 0
                    [Max] => 0
                    [Min] => 99999999
                    [Pool Calls] => 0
                    [Pool Attempts] => 0
                    [Pool Wait] => 0
                    [Pool Max] => 0
                    [Pool Min] => 99999999
                    [Pool Av] => 0

No delay. Backup pools often stay unset because they're never talked to after the initial testing they're alive.

[erk]

cgminer hangs when it looses Internet connection to the pool. Even when the connection comes back cgminer wont continue, it wont even respond to key presses, you have to kill the window. This has happened with several versions currently running 3.6.1 on window 7 32bit. Is there a work around, I just lost 7hrs of mining?

[P_Shep]

Do the pool stats have a 'ping' like time to them? Or are my stats not working... I'm getting this all the time:

                    [Calls] => 0
                    [Wait] => 0
                    [Max] => 0
                    [Min] => 99999999
                    [Pool Calls] => 0
                    [Pool Attempts] => 0
                    [Pool Wait] => 0
                    [Pool Max] => 0
                    [Pool Min] => 99999999
                    [Pool Av] => 0

No delay. Backup pools often stay unset because they're never talked to after the initial testing they're alive.

There's got to be some round-trip delay though? 20ms at least!
I'd like to keep an eye on ping times and swap pools if they start getting silly.

[-ck]

Do the pool stats have a 'ping' like time to them? Or are my stats not working... I'm getting this all the time:

                    [Calls] => 0
                    [Wait] => 0
                    [Max] => 0
                    [Min] => 99999999
                    [Pool Calls] => 0
                    [Pool Attempts] => 0
                    [Pool Wait] => 0
                    [Pool Max] => 0
                    [Pool Min] => 99999999
                    [Pool Av] => 0

No delay. Backup pools often stay unset because they're never talked to after the initial testing they're alive.

There's got to be some round-trip delay though? 20ms at least!
I'd like to keep an eye on ping times and swap pools if they start getting silly.

I think I've misunderstood. Are you asking if the pool stats show ping times to the pools? If so, they don't.

[P_Shep]

I think I've misunderstood. Are you asking if the pool stats show ping times to the pools? If so, they don't.

Right, a round trip time from request to response.

OK, I might add that stat. Somehow.

[ksenter]

By the way, my power blinked for a second, the UPS kept the computer on, and cgminer started saying:
 [2013-11-01 17:58:14] BAL 0 usb write err:(-7) LIBUSB_ERROR_TIMEOUT
 [2013-11-01 17:58:14] BAL0: RequestQueJob failed (err=-7 amt=0)
Guess it lost the connection to the Little Single.  So I hotplugged the usb cable to it, and it recovered.

Wasn't sure if it was supposed to be able to recover without having to unplug/replug the usb cable, so I thought I'd mention it.  But it probably had more to do with windows losing the device than cgminer, just mentioning it in case.

[aigeezer]

By the way, my power blinked for a second, the UPS kept the computer on, and cgminer started saying:
 [2013-11-01 17:58:14] BAL 0 usb write err:(-7) LIBUSB_ERROR_TIMEOUT
 [2013-11-01 17:58:14] BAL0: RequestQueJob failed (err=-7 amt=0)
Guess it lost the connection to the Little Single.  So I hotplugged the usb cable to it, and it recovered.

Wasn't sure if it was supposed to be able to recover without having to unplug/replug the usb cable, so I thought I'd mention it.  But it probably had more to do with windows losing the device than cgminer, just mentioning it in case.

My LS seems to handle such situations. It's connected to the same UPS as the 'puter.
 

[-ck]

As an aside: I've been running 3.5.1 since yesterday midday. No errors, so I'm still happy with it.

So, I've trialled the new 3 above.  Got zombies fairly quickly on all of them so terminated the tests soon after they occurred (I can leave them running longer if it would help prove anything).

Results as follows:
cgminer-tob started at 09:33, first zombie at 09:37, 2nd at 9:43. Stopped at 09:44.
cgminer-tor  started at 09:48, first zombie at 10:00, 2nd at 10:23. Stopped at 10:41.
cgminer-tow started at 10:45, first zombie AMU 32 after 2 minutes, but must have failed and auto-replugged (?)
                   as AMU 32 is now showing after AMU 33 in the table.

Logfiles here (without --debug):
 https://dl.dropboxusercontent.com/u/44240170/logfile-tob.txt
 https://dl.dropboxusercontent.com/u/44240170/logfile-tor.txt
 https://dl.dropboxusercontent.com/u/44240170/logfile-tow.txt

Currently running cgminer-tow with --debug. Just on the off-chance of capturing an error...
 

Just starting to catch up to you. "tob" crashed within seconds. That unhelpful Windows message "cgminer-nogpu has stopped working". On to "tor".

Edit: AMU7 has gone zombie about 15 minutes in to "tor". I'll try "tow" now.

Edit: "cgminer-nogpu.exe has stopped working" after about 15 seconds in "tow". Back to "tob" for a while, I guess.

Edit: "tob" has a zombie on AMU6 for this run about 20 minutes in. Last time it crashed early. Tough stuff to track down.

Edit: just noticed (2 hours in to "tob" run) that there is no LED on, even though AMU6 is still reported as zombie. That's a first, I think. Sorry - unit wasn't into socket far enough.

Weird - As I'm typing, all but one (see above) of the LEDs came on solid and the BAL fan has slowed. The display is still updating normally, it seems and everything claims to be hashing. I don't trust it though, so I'll restart it.

Edit: Several more zombies from time to time. Another interesting pattern, unique to "tob" perhaps - I usually get the "cgminer-nogpu has stopped working" crash now when I try to restart a zombie by disconnecting either just it  or its whole hub and replugging.

Ok well the crash is totally unexpected and may be some other change to the master code instead of the usb comms experiments. Anyway, here's some more...

http://ck.kolivas.org/apps/cgminer/temp/cgminer-chopwsync.exe

[aigeezer]

Ok well the crash is totally unexpected and may be some other change to the master code instead of the usb comms experiments. Anyway, here's some more...

http://ck.kolivas.org/apps/cgminer/temp/cgminer-chopwsync.exe

OK. For the heck of it, I tried "tob" with logging/debug on and it has been fine for two hours. I'll switch to chopwsync now.

Edit: Oops, three zombies after three minutes with chopwsync.

Edit: I added logging/debug to chopwsync and it found a zombie! I have to leave for 10 hours or so but I'll look at the log then. Meanwhile I'll run some other version.

[-ck]

Ok well the crash is totally unexpected and may be some other change to the master code instead of the usb comms experiments. Anyway, here's some more...

http://ck.kolivas.org/apps/cgminer/temp/cgminer-chopwsync.exe

OK. For the heck of it, I tried "tob" with logging/debug on and it has been fine for two hours. I'll switch to chopwsync now.

Edit: Oops, three zombies after three minutes with chopwsync.

Edit: I added logging/debug to chopwsync and it found a zombie! I have to leave for 10 hours or so but I'll look at the log then. Meanwhile I'll run some other version.

Well that didn't last long, thanks.

Throw that out and try:
http://ck.kolivas.org/apps/cgminer/temp/cgminer-cgtochop.exe

[aigeezer]

Ok well the crash is totally unexpected and may be some other change to the master code instead of the usb comms experiments. Anyway, here's some more...

http://ck.kolivas.org/apps/cgminer/temp/cgminer-chopwsync.exe

OK. For the heck of it, I tried "tob" with logging/debug on and it has been fine for two hours. I'll switch to chopwsync now.

Edit: Oops, three zombies after three minutes with chopwsync.

Edit: I added logging/debug to chopwsync and it found a zombie! I have to leave for 10 hours or so but I'll look at the log then. Meanwhile I'll run some other version.

Well that didn't last long, thanks.

Throw that out and try:
http://ck.kolivas.org/apps/cgminer/temp/cgminer-cgtochop.exe

OK. Running cgtochop now. I hope to have some log files for you shortly.

Would you clarify something, please? Are these tests all intended as replacements for cgminer.exe or cgminer-nogpu.exe? I've been using them as the latter, based on a file-size check from back when we started doing this, but I'm getting the sense that my 7970 has been mining too. Its contribution would be so small these days that it would just look like noise. Anyway, that could be a big confound for all my testing. Gulp.

[-ck]

Ok well the crash is totally unexpected and may be some other change to the master code instead of the usb comms experiments. Anyway, here's some more...

http://ck.kolivas.org/apps/cgminer/temp/cgminer-chopwsync.exe

OK. For the heck of it, I tried "tob" with logging/debug on and it has been fine for two hours. I'll switch to chopwsync now.

Edit: Oops, three zombies after three minutes with chopwsync.

Edit: I added logging/debug to chopwsync and it found a zombie! I have to leave for 10 hours or so but I'll look at the log then. Meanwhile I'll run some other version.

Well that didn't last long, thanks.

Throw that out and try:
http://ck.kolivas.org/apps/cgminer/temp/cgminer-cgtochop.exe

OK. Running cgtochop now. I hope to have some log files for you shortly.

Would you clarify something, please? Are these tests all intended as replacements for cgminer.exe or cgminer-nogpu.exe? I've been using them as the latter, based on a file-size check from back when we started doing this, but I'm getting the sense that my 7970 has been mining too. Its contribution would be so small these days that it would just look like noise. Anyway, that could be a big confound for all my testing. Gulp.

cgminer.exe -n
will always tell you what it's built for.
These are non-gpu builds since it's only that never ending source of pain, the AMU that we're testing here.

[dlasher]

Con,

upgraded from 3.6.2 to 3.6.6 two days ago, and have twice run into a new issue.

(5) different machines running 3.6.6 at (3) different physical locations, different carriers.

each miner set up for (6) pool entries, (3) btcguild, (2) eligius, (1) deepbit, set to "failover" mode.

I've now had (2) of them completely stall, no new work.

here's the messages:

[2013-11-01 17:10:07] Accepted 00cdfacd Diff 318/128 BAS 3 pool 0
 [2013-11-01 17:10:07] Accepted 01f1c8fc Diff 131/128 BAS 0 pool 0
 [2013-11-01 17:10:08] Accepted 01264c88 Diff 222/128 BAS 1 pool 0
 [2013-11-01 17:10:09] Network diff set to 391M
 [2013-11-01 17:10:19] Waiting for work to be available from pools.

and

[2013-11-02 01:38:48] Accepted 009d365c Diff 416/256 BAS 2 pool 0
 [2013-11-02 01:38:59] Accepted 000139c9 Diff 53.5K/256 BAS 2 pool 0
 [2013-11-02 01:39:05] Network diff set to 391M
 [2013-11-02 01:39:15] Waiting for work to be available from pools.

At that point ALL work in/out of cgminer stops..I get idle miner notifications from btcguild, but it doesn't fail down to eligius or deepbit.

For now I'm going back to 3.6.2, but wanted you to know I'd seen something new. I'm sorry I don't have better logs.

[aigeezer]

Throw that out and try:
http://ck.kolivas.org/apps/cgminer/temp/cgminer-cgtochop.exe

cgtochop finally found a zombie after 6+ hours. LEDs had been coming on from time to time, 7+/- seconds typically, but it was fine otherwise. I'll try to hot-swap the zombie.

Edit: the zombie was AMU6. It reassigned to 13 on the swap. The run continues.

Edit: the cgtochop run appeared to freeze about 6.5 hours in. All the LEDs came on at once and the BAL fan slowed to idle. The display looked normal, and the program shut down with the Q command. Of possible relevance - I had just started a lot of unrelated file download activity from another program a moment before the failure. I'll restart it if possible.

[jmc1517]

My run with cgminer-tow with --debug produced no errors after running all night and until midday today. Typical.

I've been running cgminer-cgtochop since 13:11
First zombie at 14:40 AMU 16 and error timeouts from AMU 0. Replugged at 14:43 recovered as AMU 34.
Another zombie AMU 16 again, but it's not in the same place (different hub too), so they have auto-renumbered (?)
Replugged the zombie AMU at 17:42 and it came back as AMU 35. Mining normally again.
Run stopped at 18:49 due to broadband problems (BT home hub - don't ask).

I'm continuing the run as it looks the most promising yet

First logfile here:
 https://dl.dropboxusercontent.com/u/44240170/logfile-cgtochop-1.txt