Open kaese2002 opened 8 years ago
Crash after 5 min :(
can´t confirm...still runs after first launch...5x290x
I have 2x380x and 4x480x
Got a R9 280x + rx 480 running on 16.10.3 amd drivers and win 10. Once a mining thread hang itselv resulting in a try to restart the gpu. The result was 0 sols on both. That watchdog doesnt seem to work as intended EDIT: Here is a screen of how it looks when it happens about 5-10 minutes in https://puu.sh/s3ZPb.png
now it´s happen...nearly one hour xD logfolder is empty
Gets as far as this and exits every time. genoil.exe -c zec.suprnova.cc:2142 -u mrdegriz.clare -p x -P 0 -g 0 1 2 3 -i 20 -w 64 Genoil's ZECMiner v0.6 13:42:22 main : Using OpenCL platform: AMD Accelerated Parallel Processing 13:42:22 main : gpu#0: Hawaii 13:42:22 main : gpu#1: Hawaii 13:42:22 main : gpu#2: Hawaii 13:42:22 main : gpu#3: Hawaii 13:42:23 gpu#0: compiling kernels/zec-sa.cl... 13:42:24 gpu#0: writing binary kernels/cache/zec-sa-Hawaii-64-20.bin 13:42:24 gpu#1: compiling kernels/zec-sa.cl... 13:42:26 gpu#1: writing binary kernels/cache/zec-sa-Hawaii-64-20.bin 13:42:26 gpu#2: compiling kernels/zec-sa.cl... 13:42:28 gpu#2: writing binary kernels/cache/zec-sa-Hawaii-64-20.bin 13:42:28 main : zec-sa#0: 0.0S/s zec-sa#1: 0.0S/s zec-sa#2: 0.0S/s zec-sa#3: 0.0S/s total: 0.0S/s 13:42:28 pool : Connecting to zec.suprnova.cc:2142 13:42:28 gpu#3: compiling kernels/zec-sa.cl... 13:42:28 pool : Connected! 13:42:28 pool : subscribed 13:42:28 pool : received new target: 0x00f0f0f0f0f0f0... 13:42:28 pool : authorized worker mrdegriz.clare 13:42:28 pool : subscribed 13:42:29 pool : already authorized 13:42:29 pool : received new job #20 13:42:37 main : 0.0S/s watcher armed for gpu#0
@degriz what if you try with just 1 card first? might need to only generate 1 bin file and if successful try running with 4
watchdog detects 0/s but when attempting to restart gpu it doesnt...running 0.6 with 10 min reboot script for time being
Wow i see zero :)
Restarted GPU 1 and 3 successfully but crashed some seconds later. Im confused about the assignment as you can see GPU 2(1) and 3(2) collapsed. Miner says 1 and 3.
Having loads of issues with this release. Watchdog doesn't restart miner. Miner seems to generate new bins on every restart regardless if they are present. Gpu is hit or miss on load and I think a lot has to do with it trying to generate bins every start. Overall Seems a bit more buggy than .5 for me. I'll check the logs in a bit but don't have time atm. Just wanted to give some feedback
i do like the colors though :smile:
:( :( :( :( too bad
running with 3 min kill and reboot script :face_with_head_bandage: ...at least itll keep submitting shares
heres a look at my log.
submitted and accepted 09:30:03 gpu#0: submitting solution 09:30:04 pool : received new job #000000362883e045 09:30:04 gpu#2: submitting solution 09:30:04 pool : not accepted: null (21) 09:30:04 pool : Retrying in 5 seconds... 09:30:05 main : zec-sa#0: 15.6S/s zec-sa#2: 22.3S/s zec-sa#3: 19.3S/s total: 57.2S/s 09:30:05 gpu#3: submitting solution 09:30:05 gpu#3: Mining thread quit 09:30:08 gpu#0: submitting solution 09:30:08 gpu#0: Mining thread quit 09:30:09 pool : Connecting to equihash.usa.nicehash.com:3357 09:30:09 pool : Connected! 09:30:10 main : zec-sa#0: 15.7S/s zec-sa#2: 18.9S/s zec-sa#3: 1.6S/s total: 36.3S/s 09:30:10 pool : subscribed 09:30:10 pool : authorized worker xxxxxxxxxxxxx.xxx 09:30:10 pool : received new target: 0x01e1e1e1e00000... 09:30:10 pool : received new job #000000362883f1a6 09:30:12 gpu#2: submitting solution 09:30:13 pool : submitted and accepted 09:30:15 main : 0.0S/s detected! Attempting to restart gpu#0 09:30:15 main : 0.0S/s detected! Attempting to restart gpu#3 09:30:15 main : zec-sa#0: 0.0S/s zec-sa#2: 18.1S/s zec-sa#3: 0.0S/s total: 18.1S/s 09:30:20 main : zec-sa#0: 0.0S/s zec-sa#2: 23.7S/s zec-sa#3: 0.0S/s total: 23.7S/s 09:30:24 gpu#2: submitting solution 09:30:25 pool : submitted and accepted 09:30:25 main : zec-sa#0: 0.0S/s zec-sa#2: 22.3S/s zec-sa#3: 0.0S/s total: 22.3S/s 09:30:30 main : zec-sa#0: 0.0S/s zec-sa#2: 21.9S/s zec-sa#3: 0.0S/s total: 21.9S/s 09:30:34 gpu#2: submitting solution 09:30:34 gpu#2: submitting solution 09:30:35 main : zec-sa#0: 0.0S/s zec-sa#2: 25.5S/s zec-sa#3: 0.0S/s total: 25.5S/s 09:30:35 pool : submitted and accepted 09:30:36 pool : submitted and accepted 09:30:39 gpu#2: submitting solution 09:30:40 pool : submitted and accepted 09:30:40 main : zec-sa#0: 0.0S/s zec-sa#2: 20.1S/s zec-sa#3: 0.0S/s total: 20.1S/s
looks like it has issues with rejected shares then upon reconnect it drops a card or two
another log same deal this one i tried with -k zec zec to get more sols but still running two instances on genoil give me about 40% boost in sols/s seems to only work on 290x and 390x cards maybe 290's if you see your gpu bouncing a lot it will make it run 100% gpu
09:35:47 pool : submitted and accepted
09:35:49 gpu#0: submitting solution
09:35:50 pool : submitted and accepted
09:35:51 main : zec-sa#0: 32.1S/s zec-sa#2: 30.7S/s zec-sa#3: 24.9S/s total: 87.7S/s
09:35:52 gpu#0: submitting solution
09:35:52 gpu#0: submitting solution
09:35:52 gpu#0: submitting solution
09:35:53 pool : submitted and accepted
09:35:53 pool : submitted and accepted
09:35:53 gpu#0: submitting solution
09:35:54 pool : submitted and accepted
09:35:54 pool : submitted and accepted
09:35:55 gpu#0: submitting solution
09:35:55 gpu#3: submitting solution
09:35:55 pool : submitted and accepted
09:35:56 pool : submitted and accepted
09:35:56 gpu#3: submitting solution
09:35:56 main : zec-sa#0: 30.7S/s zec-sa#2: 24.9S/s zec-sa#3: 33.7S/s total: 89.3S/s
09:35:57 pool : submitted and accepted
09:35:58 gpu#2: submitting solution
09:35:59 pool : submitted and accepted
09:35:59 gpu#0: submitting solution
09:35:59 pool : received new job #000000362884afd2
09:36:00 pool : not accepted: null (21)
09:36:00 pool : Retrying in 5 seconds...
09:36:01 gpu#3: submitting solution
09:36:01 gpu#3: Mining thread quit
09:36:01 main : zec-sa#0: 28.9S/s zec-sa#2: 28.7S/s zec-sa#3: 29.7S/s total: 87.3S/s
09:36:02 gpu#2: submitting solution
09:36:02 gpu#2: Mining thread quit
09:36:03 gpu#3: submitting solution
09:36:03 gpu#3: Mining thread quit
09:36:03 gpu#0: submitting solution
09:36:03 gpu#0: Mining thread quit
09:36:04 gpu#2: submitting solution
09:36:04 gpu#2: Mining thread quit
09:36:05 gpu#0: submitting solution
09:36:05 gpu#0: Mining thread quit
09:36:05 pool : Connecting to equihash.usa.nicehash.com:3357
09:36:05 pool : Connected!
09:36:05 pool : subscribed
09:36:06 pool : authorized worker xxxx.xxxxxx
09:36:06 pool : received new target: 0x01e1e1e1e00000...
09:36:06 pool : received new job #000000362884bf43
09:36:06 main : 0.0S/s detected! Attempting to restart gpu#3
09:36:06 main : zec-sa#0: 15.4S/s zec-sa#2: 12.4S/s zec-sa#3: 6.2S/s total: 33.9S/s
09:36:11 main : 0.0S/s detected! Attempting to restart gpu#0
09:36:11 main : 0.0S/s detected! Attempting to restart gpu#0
09:36:11 main : 0.0S/s detected! Attempting to restart gpu#2
09:36:11 main : 0.0S/s detected! Attempting to restart gpu#2
09:36:11 main : 0.0S/s detected! Attempting to restart gpu#3
09:36:11 main : zec-sa#0: 0.0S/s zec-sa#2: 0.0S/s zec-sa#3: 0.0S/s total: 0.0S/s
09:36:16 main : zec-sa#0: 0.0S/s zec-sa#2: 0.0S/s zec-sa#3: 0.0S/s total: 0.0S/s
09:36:21 main : zec-sa#0: 0.0S/s zec-sa#2: 0.0S/s zec-sa#3: 0.0S/s total: 0.0S/s
damn this -k zec zec rocks on the 480
but cannot confirm that for the 290x...runs slower...
-k zec zec kills my rx480 (4gb) down to 3sol/s. That only work on 8gb cards?
no i have a MSI RX480 4G
Tell what you want... with an established remote connection it´s working...runs since 3hrs now...
Interesting, nothing special about your bat?
not really
genoil.exe -c zec-eu1.nanopool.org:6666 -u xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx/xxxxxxxxxxxxxx/xxxxxxxxxx.de -p x -P 0 -g 0 1 2 3 4
I solved that problem with the zec zec command. one rig, 5 windows it works
its -k zec zec -c .....yes?
I changed the PowerShall script a little bit. Now it works in tandem with integrated watchdog. monitor_restart_ver2.zip
Maybe so you can leave it on overnight :)
Thanks @kdk2015 I was looking for something like this in another "issue" I've never messed around with powershell but it looks easy enough to edit.
@jch9678 I've never used powershell before. The script gives the "opportunity" to correct the situation before it restart task. If the minig will continue, it resets the counter.
Thanks @kdk2015 I didn't realize the miner can recover from something like that.
Incidentally for people that can't use powershell, another solution to this problem is here https://github.com/Genoil/ZECMiner/issues/66#issuecomment-257982420
Thanks to both of you for making these.
@jch9678 if your like me and cant use either a bat nor a powershell you might wana try using a programm that restarts programms on crash. https://www.raymond.cc/blog/keep-application-running-by-automatically-rerun-when-closed/ i found the 3rd one named "Reestart on Crash" pretty usefull. I set it up to check every 5 sec so worst case it restarts the miner after 5 secs after it crashes. Note that you must use the -f 2 flag for this as it will shutdown the miner if it detects 0sol/s. Works for me likee a charm and prob one of the beest solutions you can get without using a bat or powershell. (I cant vouch for that thing being virus free but nothing bad has happened to me while using it. at lest till now ^^)
Thanks @seitrox these will help me out with a couple other things I need.
@seitrox I don't get how you configure "you must use the -f 2 flag for this as it will shutdown the miner if it detects 0sol/s. " where is -f 2 flag?
genoil.exe -c zec.suprnova.cc:2142 -u mmm.mmm -p x -P 0 -g 0 1 2 3 -i 20 -w 64 -f 2
just add -f 2 to the end of your command
@mailming the -f flag can bee read in the readme of the 0.6 version of genoils miner. The flag set to 0 (default) will try to restart the gpu which doesnt work for me, -f 1 will exit the thread and -f 2 will shutdown the miner. That is where the restart on crash programm kicks in. It detects (if setup correctly ) that your programm has closed and then i told it to run the my miner.bat file to start it again. So if the miner ever hits 0sol/s which happeens quite often, it will then close and restart. For me thats atm the best solution for that problem. Until genoil fixed it , this will be a good solution. my bat looks like this : genoil.exe -c "pool" -u "user" -p x -i 20 -P 0 -f 2 -z 1 -g 0 1
@seitrox
Thanks, in that case I set my start.bat script simply like this will solve all problem.
del /F /Q %~d0%~p0%kernels\cache* genoil.exe -c us1-zcash.flypool.org:3333 -u xxxxxxx.%COMPUTERNAME% -p z -i 20 -w 64 -P 1 -f 2 echo "crashed" %date% %time% >> error.log timeout 5 %~d0%~p0%~n0%~x0
Note, it will first delete kernel cache, if error will log time. and timeout for 5 sec and restart process. I am testing it, hopefully it is working good now.
I am using 2xAMD RX 480 for mining and get up to 80S/s most of time at 65S/s
That looks like you're deleting and creating new binaries each time it crashes. You shouldn't need to do that.
This bat doesnt loop right ? Are you going to use restart on crash? if yes it has its own log thing but its maybe not that pleasent to look at.
@seitrox the script will loop, "%~d0%~p0%~n0%~x0" is looping.
oh im not good at bat things i know a bit off bash but thats it xD. But why do you delete the binaries each time do you have problems with them ?
I am trying to avoid cache corruption, not sure if it is necessary, but it doesn't hurt to do so.
Is there a proper cache location for zcash like ETH cache? I am suspecting something is related to that when crashing.
This ran fine with the script for a good hour ... the miner restarted and all I'm getting is 0.0 on every card and the display is very slowly. Best way to describe it is it's as if I'm watching someone type very very slowly?
@ofpcarnage Check to make sure you aren't running another instance (either CPU or GPU). That "slowness" happens to me when I'm running multiple instances.
all I'm running is a cpu miner but it does this without the cpu mining
good point with the -f 2 story, but how do i tell W7 not to bring those "gnoil.exe is no longer working" promt? the loop is triggered when i close this window...
You will need to modify windows registry, https://blogs.msdn.microsoft.com/alejacma/2011/02/18/how-to-disable-the-pop-up-that-windows-shows-when-an-app-crashes/
@kdk2015 Thanks for the modified PS script. It works really well. However, I did notice that I was moving along fine for ~30 minutes and then when the 0.0S/s bug hit, the script dropped right through the time-span comparison between the seconds of lastdate and get-date -gt 30 seconds and started counting where it had left off before. I checked all the code and it looks fine to me but I may be missing something... Have you ever experienced a reset of the counter back to 0? My sample set is really low as I haven't crashed again in over 2 hours now. ;)
19:09:51 main : zec-sa#0: 0.0S/s total: 0.0S/s
**-- WDOG. NEW Zeros detected. Start counting. -- [ 1/5]**
19:09:51 pool : Connecting to xxxxxxxxxxxxxxxxxx
19:09:51 gpu#0: loading binary kernels/cache/zec-sa-xxxxx-64-20.bin
19:09:52 pool : Connected!
19:09:52 pool : subscribed
19:09:52 pool : received new target: 0x01555555555555...
19:09:52 pool : authorized worker xxxxxx.rig1
19:09:53 pool : subscribed
19:09:53 pool : already authorized
19:09:53 pool : received new job #3b
19:09:56 main : 0.0S/s watcher armed for gpu#0
**-- WDOG. Zeros detected. Continue counting. -- [ 2/5]**
19:09:56 main : zec-sa#0: 22.0S/s total: 22.0S/s
19:10:01 main : zec-sa#0: 26.6S/s total: 26.6S/s
19:10:06 main : zec-sa#0: 26.2S/s total: 26.2S/s
19:10:11 main : zec-sa#0: 34.2S/s total: 34.2S/s
19:10:16 main : zec-sa#0: 28.8S/s total: 28.8S/s
19:10:17 gpu#0: submitting solution
19:10:17 pool : submitted and accepted
Notice the 28 minute break here... And then it starts counting at 3 instead of having seen NEW zeros and resetting the counter. And yes, I did lower it to 5 as I wasn't ever seeing the GPU restarting on its own.
19:38:31 pool : already authorized
19:38:32 main : 0.0S/s detected! Attempting to restart gpu#0
**-- WDOG. Zeros detected. Continue counting. -- [ 3/5]**
19:38:32 main : zec-sa#0: 0.0S/s total: 0.0S/s
**-- WDOG. Zeros detected. Continue counting. -- [ 4/5]**
19:38:32 pool : received new job #4d
19:38:37 main : zec-sa#0: 0.0S/s total: 0.0S/s
**-- WDOG. Zeros detected. Continue counting. -- [ 5/5]**
19:38:42 main : zec-sa#0: 0.0S/s total: 0.0S/s
**-- WDOG. Zeros detected. Continue counting. -- [ 6/5]**
**--KILLING PROCESS AND RESTARTING--**
Any thoughts?
Thx for the hints...works now fine as you can see on the Afterburner graph...
@kaese2002, can you share your bat script and loop script ... im still having issue with 0s/s with rx480? Thanks in advance
adjust the number of GPUs
:loop genoil.exe -f 2 -c zec-eu2.nanopool.org:6666 -u adress/worker/mail -p x -P 0 -g 0 1 2 3 4 timeout /t 1 /nobreak goto :loop
first run without problems...