Endless, unstoppable reboots. WHAT TO DO? HELP!

[Ramones1337]

There are something about 1000 messages of this kind from HiveBot. This means my farm does reboots per last 7 hours. Im kinda noob in this things so I wish youll help me. Thank you.

[itsZeroday]

Are they any error codes? Generally this is due to unstable overclocks.

[lafuente]

And sometimes this is due to an non configured motherboard bios (4G decode, PCI Links, etc.)

[Ailen]

Hi, do you know exactly, which setting could cause this? I have Asus H110 Pro BTC+, 4G decoding enabled, PCI on Gen2, no OC and I am still experiencing this issue with no error codes. The only thing I changed was mobo setting Restore AC power after loss.

[pho3nix3871]

You mean Asrock? I have exclusively used Asrock H110 Pro BTC+, and in my experience my rig was more stable with PCI set to AUTO, not gen2. Also make sure the PCI-E chips are sitting on the PCI-E slots correctly and all the way in, not tilted to one side as the USB cables have tension and can pull one side of the PCIE chip out of the socket. The Asrock H110 Pro BTC+ 's PCIE slots for some reason is very loose, so I had to get 3D printed PCIE locking clips (ebay) to lock the chip in place.

Also turn on HiveOS watchdog, it will say low-hash rate and say which GPU hung. I use NBminer and it shows which GPU was hung, so I take a note and take account of which GPU was hung. For example:

GPU0: 12345
GPU1:
GPU2: 1
GPU3: 2

Each time it hungs, I put a number and see if it is related to a specific card. Then I turn down the memory clock by a factor of 25Mhz and see how it works. So I go to 915Mhz for all cards to start with and then go back down, the cards that never hung, I can actually increase to 930 and even 950 later on.

Hope this helps!

I have 5 rigs with Asrock H110 Pro BTC+, and I am in the process of building 5 more with that mobo.

[Ailen]

Yes, I meant Asrock. It seems that in my case it was the memory temperature, which caused the reboots. I am using Phoenixminer and in the logs there was always some GPU with 0 MH/s before wdog restarted the miner. I found some corelation with memory temps of the GPUs, so I tried changing fans settings but it was lowering the memory clock on some cards that solved my problem. I am still trying to balance everything as the reboots sometimes ocures, but I’ll move my rig to colder place soon, so we will see.

Thank you for you help!