diehard2k9

UNRAID6.5.0 Crashing while transfering files

38 posts in this topic Last Reply

Recommended Posts

Hello, I am have crashes when transfering files from unassign to a share. below is the log file that is showing some ere with sata, I am not sure what to do

about this.

 

I have 4 WD gold 10TB 1 parity and 3 storage

I have 3 SSD 512 cache drive

My probelm doesn't happen all the time but I would notice that the UNRAID had stopped and all the files where not copied to the share.

I would use Krusader docker or windows 10 VM or the linux VM  to copy file from the unassign to a share.

I would like to solve this problem before I fully start using the system.

 

AMD EPYC with 64GB ram

One last thing if anyone know how to install the CPU drivers so it says install would you please help me out.

the last photo show the config and that the CPU says not install. I am wonder if

that is the reason that the overclock doesn't work and all core stay at 2.0 and not boost to 2.55 or more.

Thanks for reading and any help would be great

 

Apr 13 04:48:50 KRAKEN kernel: ata6.00: status: { DRDY }
Apr 13 04:48:50 KRAKEN kernel: ata6: hard resetting link
Apr 13 04:48:51 KRAKEN kernel: ata6: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
Apr 13 04:48:51 KRAKEN kernel: ata6.00: configured for UDMA/133
Apr 13 04:48:51 KRAKEN kernel: ata6: EH complete
Apr 13 04:50:07 KRAKEN kernel: ata6.00: exception Emask 0x10 SAct 0x0 SErr 0x280100 action 0x6 frozen
Apr 13 04:50:07 KRAKEN kernel: ata6.00: irq_stat 0x08000000, interface fatal error
Apr 13 04:50:07 KRAKEN kernel: ata6: SError: { UnrecovData 10B8B BadCRC }
Apr 13 04:50:07 KRAKEN kernel: ata6.00: failed command: READ DMA EXT
Apr 13 04:50:07 KRAKEN kernel: ata6.00: cmd 25/00:40:20:1c:21/00:05:00:01:00/e0 tag 7 dma 688128 in
Apr 13 04:50:07 KRAKEN kernel: res 50/00:00:1f:1c:21/00:00:00:01:00/40 Emask 0x10 (ATA bus error)
Apr 13 04:50:07 KRAKEN kernel: ata6.00: status: { DRDY }
Apr 13 04:50:07 KRAKEN kernel: ata6: hard resetting link
Apr 13 04:50:08 KRAKEN kernel: ata6: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
Apr 13 04:50:08 KRAKEN kernel: ata6.00: configured for UDMA/133
Apr 13 04:50:08 KRAKEN kernel: ata6: EH complete
Apr 13 04:54:56 KRAKEN kernel: ata6.00: exception Emask 0x10 SAct 0x0 SErr 0x280100 action 0x6 frozen
Apr 13 04:54:56 KRAKEN kernel: ata6.00: irq_stat 0x08000000, interface fatal error
Apr 13 04:54:56 KRAKEN kernel: ata6: SError: { UnrecovData 10B8B BadCRC }
Apr 13 04:54:56 KRAKEN kernel: ata6.00: failed command: READ DMA EXT
Apr 13 04:54:56 KRAKEN kernel: ata6.00: cmd 25/00:40:c8:18:d1/00:05:07:01:00/e0 tag 21 dma 688128 in
Apr 13 04:54:56 KRAKEN kernel: res 50/00:00:c7:18:d1/00:00:07:01:00/40 Emask 0x10 (ATA bus error)
Apr 13 04:54:56 KRAKEN kernel: ata6.00: status: { DRDY }
Apr 13 04:54:56 KRAKEN kernel: ata6: hard resetting link
Apr 13 04:54:57 KRAKEN kernel: ata6: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
Apr 13 04:54:57 KRAKEN kernel: ata6.00: configured for UDMA/133
Apr 13 04:54:57 KRAKEN kernel: ata6: EH complete
Apr 13 05:38:12 KRAKEN kernel: ata5.00: exception Emask 0x10 SAct 0x0 SErr 0x280100 action 0x6 frozen
Apr 13 05:38:12 KRAKEN kernel: ata5.00: irq_stat 0x08000000, interface fatal error
Apr 13 05:38:12 KRAKEN kernel: ata5: SError: { UnrecovData 10B8B BadCRC }
Apr 13 05:38:12 KRAKEN kernel: ata5.00: failed command: READ DMA EXT
Apr 13 05:38:12 KRAKEN kernel: ata5.00: cmd 25/00:40:48:e8:92/00:05:4b:01:00/e0 tag 22 dma 688128 in
Apr 13 05:38:12 KRAKEN kernel: res 50/00:00:47:e8:92/00:00:4b:01:00/40 Emask 0x10 (ATA bus error)
Apr 13 05:38:12 KRAKEN kernel: ata5.00: status: { DRDY }
Apr 13 05:38:12 KRAKEN kernel: ata5: hard resetting link
Apr 13 05:38:12 KRAKEN kernel: ata5: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
Apr 13 05:38:12 KRAKEN kernel: ata5.00: configured for UDMA/133
Apr 13 05:38:12 KRAKEN kernel: ata5: EH complete

unraid6.jpg

Share this post


Link to post

Your disk controller is losing contact with at least two disks and keeps resetting their links. So shut down and replace the cables and check the power to them. I can't tell which disks from the information you provided. If you continue to have problems post your diagnostics zip: Tools -> Diagnostics.

Share this post


Link to post
4 hours ago, John_M said:

Your disk controller is losing contact with at least two disks and keeps resetting their links. So shut down and replace the cables and check the power to them. I can't tell which disks from the information you provided. If you continue to have problems post your diagnostics zip: Tools -> Diagnostics.

here is the file you asked about

kraken-diagnostics-20180413-1150.zip

Share this post


Link to post

All your disks have UDMA CRC errors, are you using some sort of enclosure? If not either all SATA cables are bad/low quality or you have a different problem, like a bad PSU, controller, etc.

Share this post


Link to post
1 minute ago, johnnie.black said:

All your disks have UDMA CRC errors, are you using some sort of enclosure? If not either all SATA cables are bad/low quality or you have a different problem, like a bad PSU, controller, etc.

Yes I am use a icydock 5 bay enclosure with 4 hard drive

PSU is new but cheap and the sata cables are new

Share this post


Link to post
2 minutes ago, johnnie.black said:

Then most likely would be an enclosure problem.

Okay I will try to check my connection. everything is brand new

and I have used enclosure before without any problem.

I will try the cables, and the power connectors and

reseat the drives in the enclosure.

 

If that doesn't help I'll try it without the enclosure

Share this post


Link to post

Keep an eye on the UDMA CRC errors, if they keep increasing there's still a problem.

 

 

Quote

ID# ATTRIBUTE_NAME          FLAGS    VALUE WORST THRESH FAIL RAW_VALUE
...
199 UDMA_CRC_Error_Count    -O-R--   200   200   000    -    18

 

Share this post


Link to post
2 minutes ago, johnnie.black said:

Keep an eye on the UDMA CRC errors, if they keep increasing there's still a problem.

 

 

 

I'm a newbie to UNRAID but I had to try it cause it so powerful.

So there is alot to learn, thx for your help.

Share this post


Link to post
8 hours ago, diehard2k9 said:

Yes I am use a icydock 5 bay enclosure with 4 hard drive

 

How are you connecting it to your unRAID server?

 

Share this post


Link to post
Posted (edited)
4 hours ago, John_M said:

 

How are you connecting it to your unRAID server?

 

I'm using a 

  Ipass to 4 SATA, 51cm, W/51cm SB, S, 30AWG

 

that is connect to the motherboard and then to the icydock 5 drive bay

I have reseated the hard drives and changed the cable, I am still running into errors where the ata6 and 7 are frozen.

Apr 13 22:08:58 KRAKEN kernel: ata7.00: exception Emask 0x10 SAct 0x0 SErr 0x280100 action 0x6 frozen
Apr 13 22:08:58 KRAKEN kernel: ata7.00: irq_stat 0x08000000, interface fatal error
Apr 13 22:08:58 KRAKEN kernel: ata7: SError: { UnrecovData 10B8B BadCRC }
Apr 13 22:08:58 KRAKEN kernel: ata7.00: failed command: READ DMA EXT
Apr 13 22:08:58 KRAKEN kernel: ata7.00: cmd 25/00:00:28:66:3b/00:01:60:01:00/e0 tag 20 dma 131072 in
Apr 13 22:08:58 KRAKEN kernel: res 50/00:00:27:66:3b/00:00:60:01:00/40 Emask 0x10 (ATA bus error)
Apr 13 22:08:58 KRAKEN kernel: ata7.00: status: { DRDY }
Apr 13 22:08:58 KRAKEN kernel: ata7: hard resetting link
Apr 13 22:08:59 KRAKEN kernel: ata7: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
Apr 13 22:08:59 KRAKEN kernel: ata7.00: configured for UDMA/133
Apr 13 22:08:59 KRAKEN kernel: ata7: EH complete
Apr 13 22:09:16 KRAKEN kernel: ata7.00: exception Emask 0x10 SAct 0x0 SErr 0x280100 action 0x6 frozen
Apr 13 22:09:16 KRAKEN kernel: ata7.00: irq_stat 0x08000000, interface fatal error
Apr 13 22:09:16 KRAKEN kernel: ata7: SError: { UnrecovData 10B8B BadCRC }
Apr 13 22:09:16 KRAKEN kernel: ata7.00: failed command: READ DMA EXT
Apr 13 22:09:16 KRAKEN kernel: ata7.00: cmd 25/00:00:18:41:e1/00:01:77:00:00/e0 tag 12 dma 131072 in
Apr 13 22:09:16 KRAKEN kernel: res 50/00:00:17:41:e1/00:00:77:00:00/40 Emask 0x10 (ATA bus error)
Apr 13 22:09:16 KRAKEN kernel: ata7.00: status: { DRDY }
Apr 13 22:09:16 KRAKEN kernel: ata7: hard resetting link
Apr 13 22:09:17 KRAKEN kernel: ata7: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
Apr 13 22:09:17 KRAKEN kernel: ata7.00: configured for UDMA/133
Apr 13 22:09:17 KRAKEN kernel: ata7: EH complete
Apr 13 22:11:41 KRAKEN kernel: ata7.00: exception Emask 0x10 SAct 0x0 SErr 0x280100 action 0x6 frozen
Apr 13 22:11:41 KRAKEN kernel: ata7.00: irq_stat 0x08000000, interface fatal error
Apr 13 22:11:41 KRAKEN kernel: ata7: SError: { UnrecovData 10B8B BadCRC }
Apr 13 22:11:41 KRAKEN kernel: ata7.00: failed command: READ DMA EXT
Apr 13 22:11:41 KRAKEN kernel: ata7.00: cmd 25/00:00:40:91:0d/00:01:72:00:00/e0 tag 0 dma 131072 in
Apr 13 22:11:41 KRAKEN kernel: res 50/00:00:3f:91:0d/00:00:72:00:00/40 Emask 0x10 (ATA bus error)
Apr 13 22:11:41 KRAKEN kernel: ata7.00: status: { DRDY }
Apr 13 22:11:41 KRAKEN kernel: ata7: hard resetting link
Apr 13 22:11:41 KRAKEN kernel: ata7: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
Apr 13 22:11:41 KRAKEN kernel: ata7.00: configured for UDMA/133
Apr 13 22:11:41 KRAKEN kernel: ata7: EH complete
Apr 13 22:11:52 KRAKEN kernel: ata7: limiting SATA link speed to 3.0 Gbps
Apr 13 22:11:52 KRAKEN kernel: ata7.00: exception Emask 0x10 SAct 0x0 SErr 0x280100 action 0x6 frozen
Apr 13 22:11:52 KRAKEN kernel: ata7.00: irq_stat 0x08000000, interface fatal error
Apr 13 22:11:52 KRAKEN kernel: ata7: SError: { UnrecovData 10B8B BadCRC }
Apr 13 22:11:52 KRAKEN kernel: ata7.00: failed command: READ DMA EXT
Apr 13 22:11:52 KRAKEN kernel: ata7.00: cmd 25/00:00:c0:04:96/00:01:25:00:00/e0 tag 18 dma 131072 in
Apr 13 22:11:52 KRAKEN kernel: res 50/00:00:bf:04:96/00:00:25:00:00/40 Emask 0x10 (ATA bus error)
Apr 13 22:11:52 KRAKEN kernel: ata7.00: status: { DRDY }
Apr 13 22:11:52 KRAKEN kernel: ata7: hard resetting link
Apr 13 22:11:52 KRAKEN kernel: ata7: SATA link up 3.0 Gbps (SStatus 123 SControl 320)
Apr 13 22:11:52 KRAKEN kernel: ata7.00: configured for UDMA/133
Apr 13 22:11:52 KRAKEN kernel: ata7: EH complete
Apr 13 22:13:25 KRAKEN kernel: ata6.00: exception Emask 0x10 SAct 0x0 SErr 0x280100 action 0x6 frozen
Apr 13 22:13:25 KRAKEN kernel: ata6.00: irq_stat 0x08000000, interface fatal error
Apr 13 22:13:25 KRAKEN kernel: ata6: SError: { UnrecovData 10B8B BadCRC }
Apr 13 22:13:25 KRAKEN kernel: ata6.00: failed command: READ DMA EXT
Apr 13 22:13:25 KRAKEN kernel: ata6.00: cmd 25/00:00:78:0a:6f/00:01:c7:01:00/e0 tag 26 dma 131072 in
Apr 13 22:13:25 KRAKEN kernel: res 50/00:00:77:0a:6f/00:00:c7:01:00/40 Emask 0x10 (ATA bus error)
Apr 13 22:13:25 KRAKEN kernel: ata6.00: status: { DRDY }
Apr 13 22:13:25 KRAKEN kernel: ata6: hard resetting link
Apr 13 22:13:25 KRAKEN kernel: ata6: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
Apr 13 22:13:25 KRAKEN kernel: ata6.00: configured for UDMA/133
Apr 13 22:13:25 KRAKEN kernel: ata6: EH complete
Apr 13 22:32:02 KRAKEN kernel: ata6.00: exception Emask 0x10 SAct 0x0 SErr 0x280100 action 0x6 frozen
Apr 13 22:32:02 KRAKEN kernel: ata6.00: irq_stat 0x08000000, interface fatal error
Apr 13 22:32:02 KRAKEN kernel: ata6: SError: { UnrecovData 10B8B BadCRC }
Apr 13 22:32:02 KRAKEN kernel: ata6.00: failed command: READ DMA EXT
Apr 13 22:32:02 KRAKEN kernel: ata6.00: cmd 25/00:00:d8:0c:81/00:01:8b:02:00/e0 tag 12 dma 131072 in
Apr 13 22:32:02 KRAKEN kernel: res 50/00:00:d7:0c:81/00:00:8b:02:00/40 Emask 0x10 (ATA bus error)
Apr 13 22:32:02 KRAKEN kernel: ata6.00: status: { DRDY }
Apr 13 22:32:02 KRAKEN kernel: ata6: hard resetting link
Apr 13 22:32:02 KRAKEN kernel: ata6: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
Apr 13 22:32:02 KRAKEN kernel: ata6.00: configured for UDMA/133
Apr 13 22:32:02 KRAKEN kernel: ata6: EH complete

 

Only other thing I'm thinking to try is to connect it to the other cable and remove the drives from the drive bay and see that it is not the drive bay

everything is new.

 

image.png.2386348eb5746bc873710efee681a050.png

So I tried to copy some files and it crashed again, with the above error.

 

 

 

kraken-diagnostics-20180413-2242.zip

Edited by diehard2k9

Share this post


Link to post
Posted (edited)

Seems it is first case about AMD EPYC onboard SATA controller got such problem, I use Ryzen and only 1 disk attach to onboard and never got such problem. ( direct connect and no any enclosure ).

 

I think 3 thing one by one you may try, (1) disable all VM (IOMMU), (2) tweak something on MB BIOS, (3) bypass icydock.

 

The bad was those UDMA error count won't clear on those disk and I think at the end you may need buy another disk controller.

Edited by Benson

Share this post


Link to post
On 4/13/2018 at 11:35 PM, Benson said:

Seems it is first case about AMD EPYC onboard SATA controller got such problem, I use Ryzen and only 1 disk attach to onboard and never got such problem. ( direct connect and no any enclosure ).

 

I think 3 thing one by one you may try, (1) disable all VM (IOMMU), (2) tweak something on MB BIOS, (3) bypass icydock.

 

The bad was those UDMA error count won't clear on those disk and I think at the end you may need buy another disk controller.

Great just want I need to spend more money! Yeah I was thinking that it might be the onboard SATA controller but I have no way to test it.

I got all the setting correct in the motherboard bios but I'll look again. Maybe check them from AUTO to enable or something.

I'm going to need the IOMMU when I add the video cards that I need to passthrough.

Not sure what you mean about UDMA error count wont clear. I'm not even there yet to see if I can.

 

Looks like it didn't like the USB backup drives craddle. After I stoped using it not more reboot. Copying from over the network it slow but it also didn't crash the unraid.

I moved over 1TB. I still go some errors but not as much.

 

Any other help you can think to give me would be great.

Share this post


Link to post

I mean UDMA error counter in SMART won't clear. I also got such error when I use a Marvell SATA controller card, but it never cause crash / reboot, just prompt error and cause slow down.

 

For disable VM (IOMMU), this is for tempoary troubleshoot. With a such powerful system, how could allow for defect. >:(

 

WDC_WD101KRYZ-01JPDB1_1SJ6E45Z-20180413-1150 disk1 (sdd).txt
199 UDMA_CRC_Error_Count    -O-R--   200   200   000    -    35

 

WDC_WD101KRYZ-01JPDB1_7JJYZA2C-20180413-1150 disk2 (sde).txt

199 UDMA_CRC_Error_Count    -O-R--   200   200   000    -    25

 

WDC_WD101KRYZ-01JPDB1_7JKG3JSC-20180413-1150 disk3 (sdf).txt

199 UDMA_CRC_Error_Count    -O-R--   200   200   000    -    1

 

WDC_WD101KRYZ-01JPDB1_7JKGJ2GC-20180413-1150 parity (sdc).txt

199 UDMA_CRC_Error_Count    -O-R--   200   200   000    -    18

Share this post


Link to post
Posted (edited)

It really just looks like a connection (or possibly power) problem to me. Have you tried connecting the disks directly to the motherboard to eliminate the drive bay?

 

It looks as though your motherboard provides up to 10 SATA ports, two via regular SATA connectors and eight via two Ipass connectors, directly from the SATA controllers built into the CPUs. Therefore the Ipass connectors appear to be associated one with each CPU socket. I'm not sure what happens if you have only one CPU installed but I would expect that only its associated ports will work. Your experience will confirm if that is indeed the case. I would expect a number of this motherboard's features to be unavailable if only one CPU is used, which makes your motherboard/CPU choice seem a little strange, if you'll forgive me for saying. I can't help wondering why you would buy an expensive two-socket board but only fit a P-suffix processor - it isn't as though you can add a second one later when the need arises.

 

Edited by John_M
typo

Share this post


Link to post
29 minutes ago, Benson said:

I mean UDMA error counter in SMART won't clear.

 

That doesn't matter. What matters is whether the count continues to increase. You simply have to acknowledge the current count in the GUI to cancel the warning and if the warning returns there have been further errors.

 

Share this post


Link to post
On 13/04/2018 at 5:12 PM, diehard2k9 said:

PSU is new but cheap

 

What is the PSU? There is absolutely no sense in buying a cheap one.

  • Upvote 3

Share this post


Link to post
2 hours ago, John_M said:

 

What is the PSU? There is absolutely no sense in buying a cheap one.

 

A PSU has a quite long life span - so it might be used in your next-generation server too. So it can be seen as an investment.

 

A PSU that fails can smoke all your drives - then it doesn't matter how much parity you have. Better PSU has supervision that cuts off the power output if the PSU fails to properly regulate the voltages. Better a PSU that turns off, than seeing the magic smoke - I still haven't seen anyone manage to put the magic smoke back into the chips.

 

And a cheap PSU isn't as powerful as the good power supplies - a good PSU can handle a significant overload for a short time allowing it to better handle load spikes and allowing it to better handle the load when the system boots. The difference in capacity is because high-end PSU specify their rating based on how much power they can handle without getting overheated, while low-end PSU specify the max current the components can handle.

 

Next thing is that good PSU uses high-temperature capacitors - they do not dry out as quickly, reducing the danger of the system becoming unstable after a couple of years because the PSU fails to properly filter the output voltages.

 

See the PSU as similar to the foundation when you build a house. Build the house on a bad foundation, and you may get lots of quite expensive problems later.

  • Upvote 1

Share this post


Link to post
55 minutes ago, pwm said:

while low-end PSU specify the max current the components can handle

Or outright lie as in the case of the "free" supplies that sometimes still come with cheap cases.

  • Like 1

Share this post


Link to post
7 hours ago, John_M said:

It really just looks like a connection (or possibly power) problem to me. Have you tried connecting the disks directly to the motherboard to eliminate the drive bay?

 

It looks as though your motherboard provides up to 10 SATA ports, two via regular SATA connectors and eight via two Ipass connectors, directly from the SATA controllers built into the CPUs. Therefore the Ipass connectors appear to be associated one with each CPU socket. I'm not sure what happens if you have only one CPU installed but I would expect that only its associated ports will work. Your experience will confirm if that is indeed the case. I would expect a number of this motherboard's features to be unavailable if only one CPU is used, which makes your motherboard/CPU choice seem a little strange, if you'll forgive me for saying. I can't help wondering why you would buy an expensive two-socket board but only fit a P-suffix processor - it isn't as though you can add a second one later when the need arises.

 

Hi, to tell the truth I want to buy two CPU but I got turned off but all the dealer selling this CPU, they were all brain dead to me. When I asked question I would get B**LS*IT answers and I did buy some but had to return them due to the fact I didn't want to damage the motherboard because the blue cpu tray was not included.  I looked since jan to get the correct CPU but something or another stop me from buying from that dealer.  So I just bought from newegg and it was real cheap from that single CPU and it would let me start getting the system up and running.  Yes each CPU controls SATA and I can only use the 1 port with four drive but that is enough for me to get this beast up and running. feature that wont work because of 1 CPU is NVME ports, I dont need at this time. 1 first PICE port, dont need it at this time. Everything else runs of the first CPU. I got the MB for future proof, so when I' ready I'll get 2 dual CPU, was going to go the single CPU but that MB was out of stock, I have been waiting and I didn't want to wait anymore.

Share this post


Link to post
7 hours ago, John_M said:

 

What is the PSU? There is absolutely no sense in buying a cheap one.

I bought a cheap PSU for something else. The PSU for this MB doesn't come into stock til may 9th. So I am using this cheap 1600w PSU to get the computer up and running.

I dont know why it would matter it's not going to be pulling any major power until the new one comes into stock.

Share this post


Link to post
4 hours ago, pwm said:

 

A PSU has a quite long life span - so it might be used in your next-generation server too. So it can be seen as an investment.

 

A PSU that fails can smoke all your drives - then it doesn't matter how much parity you have. Better PSU has supervision that cuts off the power output if the PSU fails to properly regulate the voltages. Better a PSU that turns off, than seeing the magic smoke - I still haven't seen anyone manage to put the magic smoke back into the chips.

 

And a cheap PSU isn't as powerful as the good power supplies - a good PSU can handle a significant overload for a short time allowing it to better handle load spikes and allowing it to better handle the load when the system boots. The difference in capacity is because high-end PSU specify their rating based on how much power they can handle without getting overheated, while low-end PSU specify the max current the components can handle.

 

Next thing is that good PSU uses high-temperature capacitors - they do not dry out as quickly, reducing the danger of the system becoming unstable after a couple of years because the PSU fails to properly filter the output voltages.

 

See the PSU as similar to the foundation when you build a house. Build the house on a bad foundation, and you may get lots of quite expensive problems later.

Yes, I am using a 1600W PSU which is more than enough to run this system with 4 drives without pushing it to that breaking point.

All files are backed up on the original system and are only being copied to this new on.  Everything is brand new but I dont know why I am getting errors and shut downs.

 

Someone said that I should just get a controller card for the hard drive and not use the onboard. I was thinking about that but trying to keep the cost down.

Share this post


Link to post

 

1 hour ago, diehard2k9 said:

I dont know why it would matter it's not going to be pulling any major power until the new one comes into stock.

 

It isn't so much about the quantity of power but the quality. Perhaps it has split rails. So, what is the PSU (as in brand and model)?

 

Have you tried excluding the drive bay?

 

It is not inconceivable that your CPU socket might have a bent pin. It's a nuisance to do, but if all else fails I would consider removing the processor and examining the socket using a magnifying glass and desk lamp. If you see any suspect pins don't try to fix them yourself but have the board repaired.

Share this post


Link to post
7 hours ago, Squid said:

Or outright lie as in the case of the "free" supplies that sometimes still come with cheap cases.


They seldom outright lie. They just "adjust" the truth a bit by ignoring the derating factor - they may specify the maximum peak power/current the PSU can supply at 20°C while keeping quiet about how much lower power the PSU can supply at 30°C or 40°C.

 

A data sheet for a component might specify that it can handle 20A at 20° linearly derating to 0A and 100°, which means that depending on the temperature you get:

At 20°C 20A

At 40°C 15A

At 60°C 10A

At 80°C 5A

At 100°C 0A

 

It makes a huge difference if a 40A power rail is designed to survive a maximum peak current 40A, or if it is designed to handle 60-80A but with 40A specified as continuous load up to 40°C ambient temperature.


When reading a data sheet it is often more interesting to look for the information that is missing than the information that is there - the missing information is what the manufacturer prefers the customer to not know about.

Share this post


Link to post
6 hours ago, John_M said:

 

 

It isn't so much about the quantity of power but the quality. Perhaps it has split rails. So, what is the PSU (as in brand and model)?

 

Have you tried excluding the drive bay?

 

It is not inconceivable that your CPU socket might have a bent pin. It's a nuisance to do, but if all else fails I would consider removing the processor and examining the socket using a magnifying glass and desk lamp. If you see any suspect pins don't try to fix them yourself but have the board repaired.

Hi this week I will try to excluded the drive bay, waiting for a cage to up in that I can connect all the hard drive directly to the MB.

This system is only using 150watts and the PSU is a cheap one from china that is used foe a mining rig, the main PSU hasn't come in yet.

Share this post


Link to post

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now


Copyright © 2005-2018 Lime Technology, Inc.
unRAID® is a registered trademark of Lime Technology, Inc.