[nzlug] Virtualization + ethernet/ping issues.

[Michael Hutchinson]

Ok, I did the strace on the PID of the ping process I'm using that is
failing against our monitored site... I can see a definite error, but
it's not clear what is causing it, and I'm no programmer, so half the
discussions I find with google don't make sense to me, and they seem
unrelated for the most part anyway.

Here goes my strace: obfuscation of client IP only (XX)

gettimeofday({1202864466, 216269}, NULL) = 0
gettimeofday({1202864466, 216412}, NULL) = 0
sendmsg(3, {msg_name(16)={sa_family=AF_INET, sin_port=htons(0),
sin_addr=inet_addr("192.168.51.XX")},
msg_iov(1)=[{"\10\0\223\5\nP\v\321RA\262G\\M\3\0\10\t\n\v\f\r\16\17\20".
.., 64}], msg_controllen=0, msg_flags=0}, 0) = 64
recvmsg(3, 0xbfffeb40, 0)               = -1 EAGAIN (Resource
temporarily unavailable)
gettimeofday({1202864467, 216392}, NULL) = 0
gettimeofday({1202864467, 216524}, NULL) = 0
sendmsg(3, {msg_name(16)={sa_family=AF_INET, sin_port=htons(0),
sin_addr=inet_addr("192.168.51.XX")},
msg_iov(1)=[{"\10\0\"\4\nP\v\322SA\262G\314M\3\0\10\t\n\v\f\r\16\17\20".
.., 64}], msg_controllen=0, msg_flags=0}, 0) = 64
recvmsg(3, 0xbfffeb40, 0)               = -1 EAGAIN (Resource
temporarily unavailable)

This should amount to two ping packets leaving and two returning.
Tcpdump definitely see's the returning packet, but in strace I see this
unfamiliar "-1 EAGAIN" message.

I have checked our swap space, amount of physical memory available, and
disk space. None of which are running out. The CPU is rather idle most
of the time on this box.

Anyone know what the next step ought to be?

Cheers,
Mike