Brand new FreeBSD user. 11.1-RC2 locked up solid.

chromatin

chromatin

As the title says, I am a brand new FreeBSD user; we are using FreeBSD as a small workup storage server.

I've installed 11.1-RC2 shortly before 11.1-RELEASE came out and was still running 11.1-RC2 (since I had to reboot I am currently upgrading).

Today one of my staff informed me that all NFS mounts are down. FreeBSD was not responding to any traffic over gigabit or 10Gbe interfaces.

IPMI on the machine worked fine, but the console is locked up both remotely and in person. No response to keyboard input (indeed, I couldn't toggle numlock on the keyboard). No new error messages were reported to the console, or in /var/log/messages. Where else can I look? This is extremely troubling that a total lockup could happen without any sort of error message. Should I install a debug kernel?

Thanks for any guidance -- am still tentatively very positive on FreeBSD

Details:
Supermicro X10DRH board with 2x E5-2603v4 and 64GB ram.
LSI SAS3008 with 24xHGST SAS drives
Onboard gigabit ixgbe
Chelsio T520-SO-CR
 
Update to the full release version first. It's possible there was a bug that has since been fixed (it was a release candidate for a reason).

Also make sure the SuperMicro and the LSI controller all have the latest firmwares.
 
Done and done, thank you.

Uptime was over a month, so may be either (a) unpredictable or (b) manifest only with long uptime or certain load patterns (possibly over time).

I am used to kernel panics and/or dumps when something like this happens in other Unixes; is there some other place I should log for a logged error, and does the kernel have any sort of watchdog capability I could enable?
 
I think there is a watchdog function but I've never actually used it. Issues like this are pretty hard to diagnose due to their unpredictable nature. In this respect I very much prefer to see something crash and burn consistently and reproducible, that's a lot easier to diagnose ;)
 
You must log in or register to reply here.
Back
Top