<<  >> (p.321)
    Author Topic: [ mining os ] nvoc  (Read 418569 times)
    urnzwy
    Newbie
    *
    Offline Offline

    Activity: 44
    Merit: 0


    View Profile
    April 25, 2018, 10:22:58 PM
     #6401

    Hey guys,

    Anyone else having an issue mining ETH (and ZCL now) and the rig crashing freezing?

    Seems that something happen on the ETH miner, watchdog thinks it lost a GPU and tries to restart 3main. But on both of my rigs, when this happens the system freezes and I have to manually restart.

    Errors I have seen -

    GPU Utilization is low: restarting 3main...

    Thread exited with code: 29

    Is there a way to disable 3main restarting without disabling watchdog? The miner would start mining again once connected, but this freezing is my issue.

    I'm mining hush temporarily with zero issues. Was on ZCL before the fork without issues as well, same settings and whatnot. And was mining ETH for about 3 weeks with no issues. Now it won't stop crashing after being up 12-24 hours each time. It's driving me nuts.

    Two separate rigs, both 13 card (1070 and 1070 Ti's)

    The only time I have ever had a rig freeze was ultimately due to OC being too high. Lowering the GPU OC by just a bit (5 or 10) fixed my issue.

    I am not sure I totally understand your question, but if you want to disable the the 3main restart, the only way to do this is to not run the watchdog at all. To do that, change this in 1bash:

    MINER_WATCHDOG="YES"

    to

    MINER_WATCHDOG="NO"



    Shouldn't be an overclock issue. Running Hush right now for two days straight no issue.

    Running 50 core and 200 mem at 80% TDP.

    Def do not want to turn off watchdog. I've ran with these settings fine for multiple coins and long periods of time. ETH was running for over a month with zero issues and now all the sudden with the same settings, same miner, different rigs crashes every 12ish hours within minutes of eachother. Even switched servers and same issue. Something else is at play here.

    Like I mentioned, my personal desktop that mines I see ethminer restart randomly exactly when the rigs go down. It doesn't make much sense.

    I had a similar issue to this a few months ago and was told it was an issue within NVOC. That when the miner switched to "donation" mode it would freeze and the solution was to switch miners. I've tried ETHMINER, GENOIL and CLAYMORE. All the same issue.

    Oh, the miner you are referring to was an older version of the DSTM miner. The dev only had one pool configured for his donations and a network issue in Europe hosed a bunch of us for hours early one morning. That was fixed in a newer DSTM miner version, several versions ago. That was not a freeze. It was just the miner trying over and over to connect to something that it could not. When the watchdog saw that the GPUs were idle, it would restart the miner a few times and ultimately the box. This went on for hours and hours and even destroyed the boot USB drive for some folks running an older nvOC version.

    Have you checked to see what the system logs (/var/logs) say? I am assuming when you say "freeze" that the entire host becomes unresponsive and has to be hard rebooted.


    What are your miner versions?
    Run the miner update script to update them and test again

    nvOC v0019-2.0 Community Release - Miner Update Script


    My miners are all up to date. Ran the update script a while back and checked this again the other day to make sure it wasn't outdated.

    So still no clue whats up. Hush is still mining away no issue
Page 320
Viewing Page: 321