Rigs often reboot

[drale]

Hi,

my friend has two rigs with 2 x 6 RX 580 sapphire nitro 8gb.
They do not work well. Average every one hour they reboot.
Sometimes reboot is immediately and sometimes it takes about 15 minutes (CPU cooler works and lights on GPUs are on for 15 minutes and then it reboots in less than minute).
Temparature on GPUs innot above 67° C.
This is "event log" error before every reboot:

The application-specific permission settings do not grant Local Launch permission for the COM Server application with CLSID
{7022A3B3-D004-4F52-AF11-E9E987FEE25F}
 and APPID
{ADA41B3C-C6FD-4A08-8CC1-D6EFDE67BE7D}
 to the user DESKTOP-6381QKA\MinerEureka-02 SID (S-1-5-21-636305016-1107015424-1352688177-1002) from address LocalHost (Using LRPC) running in the application container Unavailable SID (Unavailable). This security permission can be modified using the Component Services administrative tool.

Please help

[1713933663]

1713933663

[1713933663]

1713933663

[1713933663]

1713933663

[1713933663]

1713933663

[1713933663]

1713933663

[PRC_May]

i have same this problem before 1 month

i stop 3 gpus and test if it reboot ..if it not reboot try to run other gpu until you fine the gpu made made this problem

i change riser or some time the problem from bios ..you must change it

[Weswayahe]

Consider from your "event log", it seem like problem from windows update. You can try to revert the last update. Because some new update of windows restricted mining activity. Or you can try to clean uninstall gpu's driver (delete folder AMD include).

[tg88]

What power supply is being used?

[jmigdlc99]

Not sure if that error log is any indication of the cause of the reboot, but i suspect that you are having a problem with your miner and GPU configuration.

I would try the following:

• Reduce or disable overclocking settings
• Use a different miner other than what you are using
• Try disabling 1 GPU and see if you mine fine with just 1 GPU (to see which GPU is causing the problem)
• DDU, then reinstall drivers

[drale]

What power supply is being used?

lc power platinum series 1200w

[nongkrong]

That's happen i think because your windows system has restricted the mining activity. Try to revert new windows update (if you have updated it).  And because your error log say "The application-specific permission settings....." and "This security permission can be modified using the Component Services administrative tool", seem like the system block your running program.
 

[nsummy]

This error is pretty common and I doubt this is the cause of the reboot.  Does it gracefully reboot?  (as in go through the windows reboot process?) or does the power just suddenly shut off?  And is there a blue screen involved?

[drale]

This error is pretty common and I doubt this is the cause of the reboot.  Does it gracefully reboot?  (as in go through the windows reboot process?) or does the power just suddenly shut off?  And is there a blue screen involved?

There are two different reboots:

1. Fast, like pressing reset button or like short power failure. Then windows reboot
2. Long, system stops but cooler on CPU works and lights on GPUs is on. After 15-20 minutes inactivity windows reboots

No blue screen and no need human interventions

[nc50lc]

lc power platinum series 1200w

6x RX 580s (185W x 6) is an overload for a 1,200W PSU.
Even though that model has 100A on its +12v rail, it is still at its peak with a 1110W load (only have 90W window).
Since you're mining Ethereum, it's alright to underclock to lower the power consumption, or use only 5 GPUs for now.

And this bugs me: "my friend has two rigs with 2 x 6 RX 580 sapphire nitro 8gb"
Two rigs with 2x6 RX580, means 12GPU in each rig? No, right? so 6GPUs each.

The error is pretty normal and shouldn't cause a reboot,
both of the rigs seems out of juice that can trigger a shutdown but your BIOS was set to restart whenever the system was abruptly powered down.

[adaseb]

This is fairly common from time to time.

With my old Radeon GPUs, they could run about a month before I had to manually reboot them. The newer RX series, not so much. For me they rarely run for over a week without any freezing issues.

Basically there are many issues why.

Most common is the risers, some are just bad quality.

Another issue is your undervolt settings, you probably overclocked it way too much and not providing enough voltage.

Another is your memory type. If you got cheaper memory, it might not run as fast as other brands.

Another issue could be your PSU. You might be overloading a rail or you might have a bad PSU.

[ay-x]

Are you sure there is no problem with the voltage settings?
Do you run the system by changing the slots and control it?
my estimates are in this direction. More experienced and experienced friends will help.

[DrX]

First I would check the power usage from the wall with watt meter.

It shouldn't be over 1000w when miner is running.

LC power isn't know for best quality

and then you should try them without bios mods and oc. After you
can get them to run stable then you can start modding.

[Coffee135]

Windows 10 often downloads its updates. I have installed various applications to block updates but sometimes this happens. The updated driver may not work correctly. Also in the Claymore settings there is a cut-off by temperature. Perhaps you have a limit to 67 degrees. My rigs are restarted maximum 1 time per week.

[xiders]

try also check your HDD with Victoria or smth. similar

[drale]

lc power platinum series 1200w

6x RX 580s (185W x 6) is an overload for a 1,200W PSU.
Even though that model has 100A on its +12v rail, it is still at its peak with a 1110W load (only have 90W window).
Since you're mining Ethereum, it's alright to underclock to lower the power consumption, or use only 5 GPUs for now.

And this bugs me: "my friend has two rigs with 2 x 6 RX 580 sapphire nitro 8gb"
Two rigs with 2x6 RX580, means 12GPU in each rig? No, right? so 6GPUs each.

The error is pretty normal and shouldn't cause a reboot,
both of the rigs seems out of juice that can trigger a shutdown but your BIOS was set to restart whenever the system was abruptly powered down.

Rig 1 has 6 GPUs

Rig 2 also has 6 GPUs

two rigs are same

[drale]

I uninstalled all windows updates, AMD driver and I reinstalled AMD driver.
Same problem, it reboots after 20 minutes.

I will try now unplug GPUs to find riser problems.

[drale]

Rig was working perfectly for months and last weeks this problem with rebooting started.
So I think problem is not with voltage.

[leonix007]

One way to troubleshoot

Start the rig with single GPU without risers and a dedicated PSU

fiire up miner --> Result?

Still reboot?

If yes, problem is in your CPU system, exclude Perif like GPU's, Risers in your suspects list

you may try to check/replace, RAM, Processor,PSU or bloated cap from MB

If No, problem is with your Perif, exclude CPU system in your suspects list

you may try to check risers, PSU cables, and GPU's


 

[drale]

I tried with elimination method to identify bad riser and I found one. When I unplugged bad riser, rig was working perfectly with 5 GPUs.
So I changed new riser and rig reboot again after 20 minutes 

[Mathero11]

lc power platinum series 1200w

6x RX 580s (185W x 6) is an overload for a 1,200W PSU.
Even though that model has 100A on its +12v rail, it is still at its peak with a 1110W load (only have 90W window).
Since you're mining Ethereum, it's alright to underclock to lower the power consumption, or use only 5 GPUs for now.

And this bugs me: "my friend has two rigs with 2 x 6 RX 580 sapphire nitro 8gb"
Two rigs with 2x6 RX580, means 12GPU in each rig? No, right? so 6GPUs each.

The error is pretty normal and shouldn't cause a reboot,
both of the rigs seems out of juice that can trigger a shutdown but your BIOS was set to restart whenever the system was abruptly powered down.

I have also 6x rx 580 running on a 1000W psu lol
now mining for 180h straight

[drale]

I have new information.

It is not problem with risers at all. When I unplug any GPU, rig works perfect. So rig works perfect with 5 GPUs.
Perhaps problem is with voltage or something else.

What should I do?

[OhHaiAndy]

I have new information.

It is not problem with risers at all. When I unplug any GPU, rig works perfect. So rig works perfect with 5 GPUs.
Perhaps problem is with voltage or something else.

What should I do?

Every time I have had issues like this it was due to using an insufficient PSU. You can get a very inexpsensive 500-750w PSU + a dual PSU connector and give that a shot. Likely will get you running.

[drale]

I have new information.

It is not problem with risers at all. When I unplug any GPU, rig works perfect. So rig works perfect with 5 GPUs.
Perhaps problem is with voltage or something else.

What should I do?

Every time I have had issues like this it was due to using an insufficient PSU. You can get a very inexpsensive 500-750w PSU + a dual PSU connector and give that a shot. Likely will get you running.

Is it possible? Rig was working perfectly more than 6 months with same PSU. That PSU is on rig right now, no changes.

[OhHaiAndy]

I have new information.

It is not problem with risers at all. When I unplug any GPU, rig works perfect. So rig works perfect with 5 GPUs.
Perhaps problem is with voltage or something else.

What should I do?

Every time I have had issues like this it was due to using an insufficient PSU. You can get a very inexpsensive 500-750w PSU + a dual PSU connector and give that a shot. Likely will get you running.

Is it possible? Rig was working perfectly more than 6 months with same PSU. That PSU is on rig right now, no changes.

If they are running at full power, then that's right at the max capacity for a 1200w PSU. If they are running at lower wattage, then it should be fine. Have you swapped the PSU out with another 1200w and tried that? I have had rigs with reboot issues because of insufficient PSUs and I've had reboot issues because the PSU was dying.

If the issue isn't the riser, but removing a GPU and it's load from the system makes the rig stable, it sounds like a power issue to me, but of course there are lots of possibilities and variables.

[leonix007]

I have new information.

It is not problem with risers at all. When I unplug any GPU, rig works perfect. So rig works perfect with 5 GPUs.
Perhaps problem is with voltage or something else.

What should I do?

Well, Congrats, you already pointed out the culprit

what to do?

options:

Repair, Replace or Add

[drale]

I have new information.

It is not problem with risers at all. When I unplug any GPU, rig works perfect. So rig works perfect with 5 GPUs.
Perhaps problem is with voltage or something else.

What should I do?

Every time I have had issues like this it was due to using an insufficient PSU. You can get a very inexpsensive 500-750w PSU + a dual PSU connector and give that a shot. Likely will get you running.

Is it possible? Rig was working perfectly more than 6 months with same PSU. That PSU is on rig right now, no changes.

If they are running at full power, then that's right at the max capacity for a 1200w PSU. If they are running at lower wattage, then it should be fine. Have you swapped the PSU out with another 1200w and tried that? I have had rigs with reboot issues because of insufficient PSUs and I've had reboot issues because the PSU was dying.

If the issue isn't the riser, but removing a GPU and it's load from the system makes the rig stable, it sounds like a power issue to me, but of course there are lots of possibilities and variables.

Is there some software or some other way to test PSU?

[honestinvestor]

The same problem, the whole point was redrawing the cards, so do not strangle them to the maximum!

[drale]

The same problem, the whole point was redrawing the cards, so do not strangle them to the maximum!

How?

If you think about reducing memory clock, I tried it, same problem with often reboot.

[Mohondoa]

The problem seem like at your windows system. Because your event log say "The application-specific permission settings do not grant Local..." and "....this security permission can be modified using the Component Services administrative tool", from that clue log it seems there are some activities that are not allowed by the windows system.

[drale]

The problem seem like at your windows system. Because your event log say "The application-specific permission settings do not grant Local..." and "....this security permission can be modified using the Component Services administrative tool", from that clue log it seems there are some activities that are not allowed by the windows system.

I already uninstalled all updates, what I need to do now?

[Kronos21]

The problem seem like at your windows system. Because your event log say "The application-specific permission settings do not grant Local..." and "....this security permission can be modified using the Component Services administrative tool", from that clue log it seems there are some activities that are not allowed by the windows system.

I already uninstalled all updates, what I need to do now?

Removing updates won't help you. In such cases, I recommend that you reinstall the entire operating system. After the update, you cannot delete everything. I recommend that after reinstalling the operating system to install software that blocks the automatic updates.