[FUG-BR] No buffer space "emX"

[Alexandre Biancalana]

2011/6/22 Felipe N. Oliva <felipe em cnsolucoes.com.br>:
> Senhores, estou com o seguinte problema:
>
> FreeBSD 8.2-STABLE
>
> Trafego:
>
>      em0                 em1                ipfw0
>  Kbps in  Kbps out   Kbps in  Kbps out   Kbps in  Kbps out
> 89343.66  10984.74  12032.96  163709.1      0.00      0.00
> 73986.44   9453.57  11249.74  162776.1      0.00      0.00
> 82020.52  10619.25  11970.38  160399.4      0.00      0.00
> 79602.59  10026.42  11442.68  162006.7      0.00      0.00
> 74788.55  10141.10  12120.73  164523.2      0.00      0.00
> 81153.43  10439.89  11903.59  162613.0      0.00      0.00
> 80163.62  10575.31  11866.03  165640.7      0.00      0.00
> 80444.00  10124.54  11401.45  157874.0      0.00      0.00
> 82067.25  10113.48  11423.59  157767.8      0.00      0.00
> 84670.40  11147.81  12452.23  164365.0      0.00      0.00
> 72053.35  10117.19  11889.06  157832.0      0.00      0.00
> 77526.03  10476.41  12242.77  158083.3      0.00      0.00
> 73478.33  10274.84  11725.72  160038.8      0.00      0.00
> 81260.33  10566.81  12079.98  166000.9      0.00      0.00
> 77436.60  10688.88  12278.61  166983.2      0.00      0.00
> 74916.26  10261.82  12594.53  163375.1      0.00      0.00
>
> # netstat -m
> 58413/1827/60240 mbufs in use (current/cache/total)
> 28910/642/29552/133120 mbuf clusters in use (current/cache/total/max)
> 28909/138 mbuf+clusters out of packet secondary zone in use (current/cache)
> 26016/328/26344/131072 4k (page size) jumbo clusters in use
> (current/cache/total/max)
> 0/0/0/16640 9k jumbo clusters in use (current/cache/total/max)
> 0/0/0/8320 16k jumbo clusters in use (current/cache/total/max)
> 176489K/3052K/179542K 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
> 4 requests for I/O initiated by sendfile
> 0 calls to protocol drain routines
>
> Ja realizamos os seguintes tunings:
>
> sysctl.conf
> net.inet.tcp.recvbuf_max=524288
> net.inet.tcp.sendbuf_max=524288
> net.inet.ip.intr_queue_maxlen=512
> net.route.netisr_maxqlen=512
> net.inet.flowtable.enable=0
> kern.ipc.maxsockbuf=384000
> net.inet.tcp.sendspace=131072
> net.inet.tcp.recvspace=131072
> kern.ipc.nmbclusters=133120
> kern.ipc.nmbjumbop=131072
>
> loader.conf
> net.link.ifqmaxlen=1024
> net.isr.direct=1
> net.isr.bindthreads=1
> net.isr.numthreads=4
> net.isr.defaultqlimit=4096
>
> Porem, de vez em sempre a interface para de responder, ou melhor responde
> isso: "No buffer space available".
>
> Alguem ja passou por isso?

Quanto o servidor tem de memória ?
Poderia mandar o resultado de um vmstat -z ?