I keep getting hardware panics in a newly built unraid server. I have run memtest for 48 hours with no errors

 

The following recipe *always* crashes the machine

1) Start copying a large amount of data to the array (>500GB)

2) Simultaneously start preclearing 2 external USB 3.0 drives

 

System will crash in 1-3 hours

 

The system specs are:

Motherboard: Supermicro C2SEA/C2SEE

CPU: Inter Pentium E6600 (dual core)

Memory: Super Talent PC3-10600 1033MHz (2x2048MB)

Disks: 3x Seagate GoFlex 3TB (was external - precleared in case, then placed in server)

UBS 3.0 card: Rocketfish USB 3.0 PCI-E (RF-P2USB3)

Case: Azza Solano 1000

 

Can you recommend strategies for isolating this, or BIOS settings that may help?

 

Thanks

syslog.txt

 

I keep getting hardware panics in a newly built unraid server. I have run memtest for 48 hours with no errors

 

The following recipe *always* crashes the machine

1) Start copying a large amount of data to the array (>500GB)

2) Simultaneously start preclearing 2 external USB 3.0 drives

 

System will crash in 1-3 hours

 

The system specs are:

Motherboard: Supermicro C2SEA/C2SEE

CPU: Inter Pentium E6600 (dual core)

Memory: Super Talent PC3-10600 1033MHz (2x2048MB)

Disks: 3x Seagate GoFlex 3TB (was external - precleared in case, then placed in server)

UBS 3.0 card: Rocketfish USB 3.0 PCI-E (RF-P2USB3)

Case: Azza Solano 1000

 

Can you recommend strategies for isolating this, or BIOS settings that may help?

 

Thanks

you are running out of ram.  Use the options in preclear_disk to limit the block size used in reading and writing the disk.

 

      -w size  = write block size in bytes

 

      -r size  = read block size in bytes

 

      -b count = number of blocks to read at a time

 

by default, the number of blocks to read at a time is 200

 

By default, the write block size is 2048k

 

The read block size, by default is an entire cylinder (as defined by the "Units" output of fdisk -l /dev/sdX ). 

root@Tower2:/boot# fdisk -l /dev/sdc

 

Disk /dev/sdc: 2000.4 GB, 2000398934016 bytes

1 heads, 63 sectors/track, 62016336 cylinders

Units = cylinders of 63 * 512 = 32256 bytes

 

The default number of blocks to read at a time is 200.

for the above disk, in the absence of a  -r, or -b option I would be reading in 32256 * 200 bytes at a time ( 6,451,200 bytes )

 

I would try something like:

preclear_disk.sh -b 10 -r  16384 -w 16384 /dev/sdX

 

The pre-clear's will run slower... but odds are they will not run out of memory.

oooooo  you probably have it! After the sole preclear finishes, I will give it a try

 

Many thanks