Watchdog reboot the Rig

#1

Hello,
I am new in this site, so excuse me if I make a lot Mistakes.
I test this super monitoring Farm site with wonderment, but after a few moments all was black and like in hell… >:)
The context :
Mother card Asrock B250 Mining Expert ith Celeron 2,9 GHZ and 16 GB of RAM
Video Card : 13 Sapphire Radeon RX 8GB
All cards are overclocked and work good on windows 10 with MHs between 26 to 29.
All cards work correctly with Xubuntu with MHS betweens 14 to 18… (not good but work…)

And with HiveOs, All cards work a few minutes… euhhhhh not now a few secondes with MHs like windows 10 (Youppy) and reboot the rig and just 5 cards after reboot are working again with MHs between 2 to 7… a Big shit !

Why so much hatred… :’(

Look my beautifful picture before reboot, with temp 511°, like I Said, here is the hell ! >:)

Could you help me ?

Thank a lot per advance

Christophe

#2

If card reported 511 degree it’s mean she off-the-bus.
In most cases it’s over-overclock or huge power draw from the wall and as result power shortage. Of course there may be another trivial reasons.

What PSU do you use?
Can you add screenshot with OC profile ?

#3

Hi HaloGenius,
Thank’s for your response.

For PSU : I have 3 corsairs RM 750 X connected => 2150 W
These cards use normaly a max of 120 W then less (1560W) in full power.

For the Overclocking, I change the Vbios to recopy the Memory config value of 1500Mhz to the upper frequenzies, that’s all. And cards work a long time correctly on windows and Xubuntu.

For the OC in HiveOs, I have test with 1250 Mhz for the GPU and 1900Mhz for the Memory (Like on Windows with afterburner) but after the first crash I delete all to come back in the default.

For information, after the reboot, 9 PCI slots are in error and I must stop the rig to refind 13 slots active.

That is my bad history

Christophe

#4

What core voltage are you using?
I’ll ask you again to provide a screenshot with AMD OC settings

#5

The info :

Below the core voltage, not changed :

And below the frequenzies changed :
For Samsung

For Hynix

No more, I hope to have respond to your request.

I can send you the Vbios file, all before and after flashed are saved.

Br

#6

I’m ask for AMD OC settings in HiveOS (rig control panel from web) but you got me memory timings from VBIOS . It’s not the same thing.

So, once again - publish your AMD OC settings from rig web control panel

#7

Ahhhhhhhhhhh OK this… but as I said, just 80% of FAN but no else…
Below th OC attached in my Rig Profil :

But before I had set parameters like :
Core Clock : 1250 Mhz
Memory Clock : 1900 MHz

But I said allready this… I am old, it’s the reason that I repeat myself

It’s All, nothing other…

Do you have specific parameters for my pissue ?

Br

#8

God damn it … We lost miner )))

The situation is gradually clearing. To the above screen I’ll ask you to add one more - give me screenshot with parameters of miner (you can cut screenshot without your wallets addresses - they do not need me), only the parameters

#9

Cool, always in the fog for me…

Normaly I want mining just with claymore in the eth-eu1.nanopool.org:9999

But in the first time I mined with Sia too, …And I don’t know how I have disable it…

Thank again for your help

#10

Hi HaloGenius,
I made a lot of tests, in the order :
First : I delete all OC config and reload HiveOS => Same Error all card after 5 sec were KO
Second : I remade the key boot, like for the first time, => the card work 1 minutes between 2 or 7 Mh/s… shittt Mh/s… and 2 card in 25 / 26 Mh/s, good, but 2 of 13… and after all cards were KO like in the first test…
I don’t know if I must recreate a new profil or what…
Ok for the moment the cards work with Xubuntu at 18 Mh/s… it is not the best but 13 of 13 work very stable…
Best regards and good sunday.
Christophe

#11

Hi,
I close my issue after a lot of test I found a stable configuration.

In the first time I decrease the number of cards to 7 and the rig works good.
After I add each time 2 cards until 11 cards, and works always good.

Then, my conclusion is like you thinked the first time a PSU issue. I think but not toal sure than to many cards were in the same PSU.
Now I am careful to dispatch cards correctly in the different PSU.

I test this night with 13 cards and hope have no more issue.

Thank’s a lot for your help

Christophe