Fatal trap 12: page fault while in kernel mode

[JayG30]

System has been running 24/7 on the same hardware for the past 2 years without issue. Just upgraded last week from 11.0-U1 to 11.1-U4. Everything seemed fine. I was messing around with Docker/Rancher per the FreeNAS manual and have had that up and running for only a few days. This morning I show up to work and it is down. Connect to the server via IPMI and see this warning;

Code:

Fatal trap 12: page fault while in kernel mode
cpuid = 22; apic id = 26
fault virtual address	= 0x18
instruction pointer	  = supervisor write data, page not present
instruction pointer	  = 0x20:0xffffffff80a29142
stack pointer			= 0x20:0xffffffffe27fa5828e0
frame pointer			= 0x20:0xffffffffe27fa582900
code segment			 = base 0x0, limit 0xfffff, type 0x1b
						 = DPL 0, pres 1, long 1, def32 0, gran 1
processor eflags		 = interrrupt enabled, resume, IOPL = 0
current processor		= 12 (swi4: clock (18))

Specs on the box since I know people love to see that stuff;

Show : Hardware Specs

Barebones Server: Intel R2312GL4GS
Motherboard: Intel S2600GL
CPU: 2 x Intel Xeon E5-2670
RAM: 16 x 8GB (128GB) PC3L-10600R/DDR3-1333 ECC (Nanya NT8GC72C4NG0NL-CG)
Storage Disks: Seagate 4TB NAS HDD (ST4000VN000)
ZIL: Intel SSD DC S3710 400GB
Boot disk: 2 x 32GB SanDisk Ultra Fit
Note: run the 4 Intel nics in a LAGG

Something makes me thing this was caused by the Rancher VM I had running. I know most of the time people like to point to "hardware" but I doubt it. All Server grade gear that has been running fine for years. But I'm always hesitant to report a bug. I'm now running it with the Rancher VM removed and no VM's currently running. Will be curious to see if the issue arises again in the next few weeks.

 

[dlavigne]

Has this happened again? If it does, please report it at bugs.freenas.org and attach a debug to your report.

 

[JayG30]

Has this happened again? If it does, please report it at bugs.freenas.org and attach a debug to your report.

Since removing the Docker/Rancher stuff (as outlined HERE) I have NOT experienced the fatal error. So what I've experienced is this;

• 2yr+ through multiple FreeNAS versions and never had a fatal error
• ~2days with Docker/Rancher and fatal error
• ~1wk after removing Docker/Rancher and no fatal error

Obviously can't make a definitive conclusion based on a small sample set but it sure feels like something wrong somewhere in the virtualization stack.