[Bug 235856] FreeBSD freezes on AWS EC2 t3 machines

[bugzilla-noreply at freebsd.org]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=235856

--- Comment #8 from Charles O'Donnell <cao at bus.net> ---
Same symptoms still occurring regularly here. Begins with "nvmeX: Missing
interrupt" in the logs, which can go on for a few hours, then the server
becomes unresponsive. Typically occurs 4-6 days after boot. Only the most
heavily loaded production servers are affected. Other servers have never
exhibited this behavior. Have tried a variety of instance types. At this point
I have it monitored so I can do a graceful cycle before it locks up. 

See also bug #236989

Uname:

FreeBSD 12.0-RELEASE-p10 FreeBSD 12.0-RELEASE-p10 GENERIC  amd64

Log excerpt:

Aug 27 03:02:45 host kernel: nvme4: Missing interrupt
Aug 27 03:02:56 host kernel: nvme1: Missing interrupt
Aug 27 04:19:11 host kernel: nvme1: nvme4: Missing interrupt
Aug 27 04:19:11 host kernel: Missing interrupt
Aug 27 04:19:30 host kernel: nvme0: Missing interrupt
Aug 27 05:16:58 host kernel: nvme1: Missing interrupt
Aug 27 05:16:58 host kernel: nvme4: Missing interrupt
Aug 27 14:58:13 host kernel: ena0: Keep alive watchdog timeout.
Aug 27 14:58:13 host kernel: ena0: Trigger reset is on
Aug 27 14:58:13 host kernel: ena0: device is going DOWN
Aug 27 14:58:13 host kernel: nvme4: Missing interrupt
Aug 27 14:58:13 host kernel: nvme1: Missing interrupt
Aug 27 14:58:16 host dhclient[417]: send_packet6: Network is down
Aug 27 14:58:16 host dhclient[417]: dhc6: send_packet6() sent -1 of 52 bytes
Aug 27 14:58:17 host kernel: ena0: free uncompleted tx mbuf qid 5 idx 0x25c
Aug 27 14:58:17 host kernel: ena0: ena0: device is going UP
Aug 27 14:58:17 host kernel: link is UP

-- 
You are receiving this mail because:
You are the assignee for the bug.