Rig total freeze, requires power reset

Hello .
if it can help…
I changed , on my 2080 TI from ethminer to phenixminer… and now, no more card sleeping :slight_smile:

I have been facing the same issue, I also have a 120GB Kingston SSD connected to Server PSU.

I ruled out everything except for SSD. I read somewhere SSDs are more prone to crash than Flash Drives which is absurd.

I have also tried an M2 SSD vs. a brand new Samsung 128 GB USB vs. a brand new SanDisk 32GB USB vs. an older 8 GB USB 3.0.
The freeze happens on all of them.

I tried GMiner 2.54 vs 2.49 and T-Rex 20.3 vs 20.1. The longest runs I had with T-Rex 20.1. But still a freeze after a day with + absolut core clock function + Hive 0.6-203 + 455.45. Driver so it should not be the Miner as a root cause.

I have also checked the downgrade to 06.190 after 1 day I had again the freeze.

So I start to believe that @juanenk is right with the riser card thesis.
I have risers in Version 008S, which were new and only 3 month old that’s why I was initially not suspicious about.
@juanenk, which one did you buy that solved the problem on your system ?

At the moment I am reducing OC on memory after each crash by 20 MHz.
I started at 2600 and I am already at 2560 after 2 crashes. Let’s see how it goes and where I end up with my new “stable” OC.

Well mine where new too and they were the problem. I bought them in february in aliexpress and the were suppossed to be VER 009S.
Now I’m using “BEYIMEI PCI-E 1x a 16x GPU Riser– VER010-X” bought in amazon and they work flawlessly. Non stop since I post the last message.

Maybe you can try to pinpoint the defective one by replacing them one by one and letting the rig work until it crashes.

Update:

I have on all GPUs the new raisers since 2 days “Beyimei… V010-X” → still the same issue.
Now I changed the USB Slot, coz it felt a kind of hot → still the same issue.
I even tried SMOS to check if it is an issue with HiveOS → still the same issue.

So it seems to freeze due to CPU or Mainboard issue.
Also the log file just writes many “@” as last logs.
There is not much more left as hardware components.

Now I changed one more time the USB Stick and I changed BIOS settings with a slight OC on the CPU as it was also mentioned here earlier. Let’s see the results.

Keep you posted.

Hi guys,

Exactly the same issue here.

Asus B450-F Gaming II (but also tried msi gaming plus, and Asus b450-f gaming)
Ryzen 1600 (also tried athlon 3000g)
Asus rtx 3070 x6
Crucial ballisstix ddr4 (also tried other ram)
Kingston 120GB ssd (but also tried running the os from USB)

Tried replacing the risers, no luck (although I’ll try again)
Tried virgin OC, over clock, underclock
Tried locked core voltage at 1000-1100
Tried changing bios settings, flashing bios
Upgraded to newest hiveos (current)

Wanted to bring your attention to something I noticed today - while the keyboard and mouse is unresponsive, I can see that the miner has crashed but the hiveos is still showing accurate time. Can anyone else confirm this finding?

Also, I can see that everyone posted they’re using a B450 chipset, is this correct?
Also, seems that in each rig there was at least one 3060ti or 3070?

Trying to figure out if this is a gpu or motherboard/amd fault

Hi Matt,

After following bios change my OS is running stable without a freeze since 2 days.
May be you can try as well following changes in the BIOS:

That is the only thing that helped me so far.
Keep you posted if continues to go well…

1 Like

Thanks mate, will try tomorrow. Tonight I’ve hot a day off from the mining frustration. But seriously - anyone has any ideas, even stupid or scientifically implausible ones, please share, I’ll try anything!!!

Also, I’ve asked the question to hiveos support and shared a link to this thread - at least they’re aware that this is not an isolated matter.

He suggested changing a miner - tried the ethash and experienced a freeze so it’s not that.

Another suggestion was to run a single card to identify a problem one, add another card, etc. If the bios spec won’t work, that’s my next step.

Someone else suggested it’s probably a RAM crash but I highly doubt it, somehow I think that if it was ram, the os clock wouldn’t be running. Still, will try running just one chip, then another, can’t be that both are faulty.

In terms of further narrowing the fault - is (or was) anyone here reporting this issue on an Intel architecture?

Having same issues as everyone, tried everything with new risers, cables etc. My rig will sometimes freeze between 4-10 hrs. I have rtx 3060ti and rtx 3070 in the rig. Motherboard is Asrock b450 pro4. I’m starting to believe is hiveos and amd motherboard compability issues.

Welcome to the club!!

Hiveos support is trying to replicate the issue and asked to provide them with as much data we can. There’s a bunch of us so it won’t be a problem of creating a nice diagnostic dataset.

In order to get the best data set possible, can we all please:

  1. When posting here for the first time take a screenshot of your config.

  2. Screenshot of your GPUs

3.Screenshot of your OC settings

  1. Run the “htop” command and keep it on at all times. Once the rig froze, pls take a pic of the display

I’m going to create a spreadsheet to put it all in one place

1 Like

Ah… Just noticed that there are limits to replies and embedded graphics

Pls send me the screenshots on telegram @matt_le_sands, Telegram: Contact @Matt_le_Sands

Hi,
Same issue here since Sunday, what a mess! 2 days and nights to look for something…
MB : Asrock H110 Pro
GC : 8 x 3070, 2 x 3060 TI

I’ll send you my screenshots.

Good luck!

Hi,

To execute “htop” :

Send it to Telegram: Contact @Matt_le_Sands

Bye and good luck

Hi guys,

My rig has been on for 2d 12hrs, no freezes. I haven’t done anything other than changing miner from T-Rex to ethminer.

Will keep updated.

Has anyone experienced crashes recently?

Hi mate!
Quick feedback from my side.
I made a deep investigation on my rig (Asrock H110, 3 x 3060 TI, 7 x 3070, SSD) and I have made the following change.

  • Only one power cable for 2 GPUs
  • Only one SATA power cable for 2 risers
  • Change some riser (some new one didn’t work…)

Since no more freeze but sometimes (around 2 per day) I have GMINER who restart due to a PCI fail.
This issue has been identified when I used the following RUN COMMAND grep -a Xid /var/log/syslog | tail -n 10.
Identify the fail riser still in progress.

I hope this will help you.

Bye

Hey guys, short update from my side.

After trying several changes:
New risers - no improvement
New usb sticks - no improvement
Run on SSD - no improvement
Switch between different miners (gminer and t rex) - no improvement
Downgrade to Hive 6.190 - no improvement
Try another OS - no improvement
Reduce significant the OC - no improvement

The only thing that helped and where my rig is running since 6 days without crashes is the BIOS modifications as shown in the picture a bit higher.

I propose to try the bios changes first before changing Components since it seems to be a specific AMD / B450 problem and since it is the easiest change with least effort.

Good luck and keep you guys updated if anything changes on my rig.

1 Like

Guys, I will repeat the request from Support - please launch htop and post screenshots of htop once the rig is frozen so this can be looked at

1 Like

So I tought this issue was gone (I had like 3 weeks stable), and now it’s back. I did nothing… It seems so random

Oh guys. It is a shame that so many of us have same/similar issue. However I am glad that community cares and shares their experience. Mining is not plug and play as gamers like to call it

Never ever use SATA to power your riser… unless you want to burn your house down

PCI SIG specifications say 75W max power draw over the PCIe slot. you’re doing 2 so that’s potentially 150W

A Sata cable/connector is rated for 54W! (GPU Mining Resources: Maximum Safe Wattage of PSU Cables

good luck!