Idle Igb Mbuf Usage

[mkubilus]

Using FreeBSD 8.1 on AMD64. I have several systems that are showing a high mbuf usage even while not doing anything with the network.

# netstat -m
8194/1856/10050 mbufs in use (current/cache/total)
4096/1238/5334/25600 mbuf clusters in use (current/cache/total/max)
4096/766 mbuf+clusters out of packet secondary zone in use (current/cache)
0/595/595/12800 4k (page size) jumbo clusters in use (current/cache/total/max)
0/0/0/6400 9k jumbo clusters in use (current/cache/total/max)
0/0/0/3200 16k jumbo clusters in use (current/cache/total/max)
10240K/5320K/15560K bytes allocated to network (current/cache/total)
0/0/0 requests for mbufs denied (mbufs/clusters/mbuf+clusters)
0/0/0 requests for jumbo clusters denied (4k/9k/16k)
0/0/0 sfbufs in use (current/peak/max)
0 requests for sfbufs denied
0 requests for sfbufs delayed
0 requests for I/O initiated by sendfile
0 calls to protocol drain routines

These systems *always* show a minimum mbuf clusters in use of 4096 and when under any kind of network load they quickly run out of mbufs and become unresponsive on the network.

I've even tried doing a /etc/rc.d/netif stop and I show the same results with netstat -m

The systems all are using the igb network interface. Other systems on the same OS that are do not have igb interfaces do not show this high idle mbuf usage.

Any insight is appreciated,
Matt

 

[mkubilus]

#  vmstat -z | grep mbuf
mbuf_packet:              256,        0,     4096,      760, 137626943,        0
mbuf:                     256,        0,     4108,     1041, 146029247,        0
mbuf_cluster:            2048,    25600,     4856,      644,   201605,        0
mbuf_jumbo_page:         4096,    12800,        0,      612,  1601745,        0
mbuf_jumbo_9k:           9216,     6400,        0,        0,        0,        0
mbuf_jumbo_16k:         16384,     3200,        0,        0,        0,        0
mbuf_ext_refcnt:            4,        0,        0,        0,        0,        0