Kernel panics related to vm_page

[short-stack]

I had upgraded to Corral and needed to rebuild my zpool, so I copied all of my data off of my NAS and rebuilt the server from scratch.
Now when I go to re-import my data, I can almost trigger a kernel panic on queue.

I re-installed 9.10 this afternoon, and I've had a panic over 4 times now. It's generally when I'm trying to rsync or do a cp -R to move my files back in the datastores where they belong. I have multiple TB of data back in place now, but still working on the last ~300 GB. But something the copying of file is making the virtual memory angry and then kernel panic.

I enabled autotune to see if that would help anything, but thus far it has not.

I've created a bug report with debug file #23199 but figured I would put it here in case anyone had any ideas.

 

[Robert Trevellyan]

Have you scrubbed the pools? Maybe one of them is corrupted.

 

[short-stack]

Have you scrubbed the pools? Maybe one of them is corrupted.

scan: scrub repaired 0 in 3h28m with 0 errors on Sat Apr 8 15:39:43 2017

I also ran memtest overnight, and I know there might be issues because I am forced to use non-ECC RAM because of my motherboard, but I noticed that my RAM timings were off. I had updated my BIOS firmware, and it reset all of my settings so I changed the timings to the correct values and set the clock to the proper speed also.

Don't want to jinx it, but it's been up for 14 hours now! I'm used to my old FreeBSD servers being up for years hah.