[Bug 205706] Watchdog timeout on em driver under heavy traffic on a bridge configuration

[bugzilla-noreply at freebsd.org]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=205706

daulton at tetmail.ca changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |daulton at tetmail.ca

--- Comment #10 from daulton at tetmail.ca ---
I receive the same problem except on FreeNAS-11.3-U2. The problem started a few
updates ago, sometime around 11.2-U6. I have been using the same system since
FreeNAS 9.10 and it did not have this problem back then. The motherboard is a
Supermicro X8SIA which has dual onboard Intel 82574L Gigabit Ethernet
Controllers.

I noticed from my Zabbix logs this issue mostly seems to happen during or after
my backups run to the system, which occur over NFS. This is a lot of sustained
traffic from two hypervisors backing up all guest hosts.

In the logs I get the same sort of error others shared:

Apr 18 19:18:42 freenas em0: Watchdog timeout Queue[0]-- resetting
Apr 18 19:18:42 freenas Interface is RUNNING and ACTIVE
Apr 18 19:18:42 freenas em0: TX Queue 0 ------
Apr 18 19:18:42 freenas em0: hw tdh = 0, hw tdt = 905
Apr 18 19:18:42 freenas em0: Tx Queue Status = -2147483648
Apr 18 19:18:42 freenas em0: TX descriptors avail = 119
Apr 18 19:18:42 freenas em0: Tx Descriptors avail failure = 0
Apr 18 19:18:42 freenas em0: RX Queue 0 ------
Apr 18 19:18:42 freenas em0: hw rdh = 0, hw rdt = 1023
Apr 18 19:18:42 freenas em0: RX discarded packets = 0
Apr 18 19:18:42 freenas em0: RX Next to Check = 0
Apr 18 19:18:42 freenas em0: RX Next to Refresh = 0

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