Fatal trap 12 - while copying data from server to workstation

[gpsguy]

Last night, I received the dreaded Fatal trap 12: page fault whie in kernel mode, while trying to copy (via Robocopy) ~180Gb of data (CIFS share) from my server to a eSATA connected disk on my pc.

Both machines have Gb NIC's, connected to a Gb switch. I disabled my antivirus and everything else in the system tray, before starting the copy.

When the problem started, I couldn't copy more than 4 minutes worth of data (~40Mb/sec) to my eSATA disk, before the server crashed.

I did some searching here and saw that running memtest86 was the recommended first step. I ran it for ~8hours last night. It ran 3 passes without any errors.

This morning, I decided to retry the copy. Instead of copying it to eSATA, I copied it to the SATA drive in my pc. This worked much better, but the server still froze twice. Here's a timeline of events.

Code:

 
 7:47am - begin copy
 8:18am - still running
 8:36am - server froze  (cumulative transfer = 93Gb)
 9:28am - reboot & restart copy
 9:33am - server froze  (cumulative transfer = 102Gb)
 9:34am - reboot & restart copy
10:14am - done  (cumulative transfer = 182Gb)

FreeNAS-8.0.4-RELEASE-x64 (10351)
Mobo: Zotac M880G-ITX w/Dual Core AMD Turion II Neo K625
RAM: 8Gb Crucial DDR3 (used Crucial's configurator, to select 8Gb kit)
Hard disks: 2 x 2Gb Seagate Barracuda XT (ZFS mirror)
Power: SeaSonic X SS-400FL
Flash drive: 8Gb Patriot Rage XT

Attached is a photo of the error. Bill

 

[peterh]

Bad memory, or some other bad hardware or improper bios setup.