Random crashes and now degraded pool?

[Kristan McDonald]

Hi all,

Have been running Freenas for many years and currently running 9.10.2-U1.

After realising the Virtualbox jail was broken over Christmas I proceeded to make my own so I could run my VMs again. All happy, life was good. However after the last U1 update, I *think* running something in VirtualBox is resulting in a panic somewhere. If I leave my box up, without starting a VM, it seems ok. However if I run one of the VMs, after a period of time (many hours) I find my FreeNAS box has rebooted and is at the console telling me one of my drives is dirty (I have a UFS SSD mounted for keeping the VMs on). I fsck it, reboot and the box comes up again.

Weekly scrub started last night however and now my array (6 disks, two raidz1's of 3 each) is degraded with each disk showing many between 10 and 28 checksum errors and potentially some dataloss (fortunately not something vital).

So two questions -
1. Would something running in a jail be able to bring down the whole box? Happy to find another solution if it means I have a stable system.
2. Where do I start looking for the cause of the crash? quick read of messages has nothing, nothing useful in last, no crashdumps..

Any advice gratefully appreciated!

 

[dlavigne]

It shouldn't, assuming the hardware specs are up to the task. What are your hardware specs?

 

[Kristan McDonald]

Hardware should be more than capable. Supermicro X8DTH-iF with dual X5570's and 48GB ECC. Drives are on a flashed LSI 9211-8i. Nothing weird or non-standard really.

 

[dlavigne]

If you don't get any suggestions here, it's probably worth creating a bug report that includes a debug (from System -> Advanced -> Save Debug). If you end up creating a bug report, post the issue number here.

 

[Kristan McDonald]

Ok so it's not running VM's. Didn't start anything this afternoon, another crash. Bug reported 20310. Going to revert to 9.10.2 in a few days if I can't get to the bottom of it.