I've encountered what I think is a strange issue. I'll spare you the
diagnostics leading up to finding the cause, but I've determined the
following:

CentOS 6.5, kernel 3.4.41 or 3.4.86, NET-SNMP 5.5 (from CentOS6.5 repo)
ixgbe 3.19.1 / 3.21.2: Once loaded, net-snmp stops responding, and normal
ioctls (ifconfig, ip addr, or my own code, I think just SIOCETHTOOL) take a
long time (1s+) to complete.

For net-snmp, specifically, if I run the snmpd daemon (simple config,
"rocommunity public / rwcommunity private") and then run anything using the
SIOCETHTOOL ioctl (like ifconfig, or "ip addr") I will periodically see a
1s delay in returning from the ioctl syscall (according to strace -T
<cmd>). If I then run "snmpset -v 2c -c private localhost
nsCacheTimeout.1.3.6.1.2.1.2.2 i 3", I will increase the likelihood of the
ioctl seeing the delay. If I further set that timeout to only 1s, net-snmp
will completely stop responding to SNMP requests (but doesn't die). If I
then remove ixgbe (rmmod ixgbe), snmp will resume functioning normally, and
all ioctl operations will return almost immediately. If I load the in-tree
driver (3.8.21-k), I do not see this issue.

I don't know where to go from here.



Thanks,

Scott Silverman | IT | Simplex Investments | 312-360-2444
230 S. LaSalle St., Suite 4-100, Chicago, IL 60604
------------------------------------------------------------------------------
_______________________________________________
E1000-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/e1000-devel
To learn more about Intel&#174; Ethernet, visit 
http://communities.intel.com/community/wired

Reply via email to