No buffer space available

[Thiago Esteves de Oliveira]

Marc,

My machine is working with the kernel that comes with 6.1-STABLE (I  archived this good kernel
before upgrade the OS to 6.2).
No, I'm not using geom.

Can you send your dmesg.boot and "sysctl -a  kern" output?


Marc G. Fournier wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
>
> Thiago ...
>
>   I'm just curious here, but are you by any chance using geom at all?  The only
> machine I have that seems to be affected like this (where netstat -m doesn't  seem to indicate a
problem with mbufs) is using gmirror ... the rest all use  hardware RAID controllers ...
>
>   Its a long shot, but so far, its the only one I seem to be able to draw :(
>
>
> - --On Sunday, April 01, 2007 17:07:08 -0300 Thiago Esteves de Oliveira 
<thiago at lamce.coppe.ufrj.edu.br> wrote:
>
>
>> I've tried to increase the kern.ipc.nmbclusters value but it worked only when I changed the
kernel to an older one.
>>
>> netstat -m (Now it's working with the same values.)
>> -
>> 515/850/1365 mbufs in use (current/cache/total)
>> 512/390/902/65024 mbuf clusters in use (current/cache/total/max) 512/243 mbuf+clusters out of
packet secondary zone in use (current/cache) 0/0/0/0 4k (page size) jumbo clusters in use
(current/cache/total/max) 0/0/0/0 9k jumbo clusters in use (current/cache/total/max)
>> 0/0/0/0 16k jumbo clusters in use (current/cache/total/max)
>> 1152K/992K/2145K 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
>> 2759 requests for I/O initiated by sendfile
>> 2982 calls to protocol drain routines
>>
>> Ethernet adapters
>> -
>> em0: <Intel(R) PRO/1000 Network Connection Version - 6.0.5> port 0xec80-0xecbf m em
0xfebe0000-0xfebfffff irq 10 at device 4.0 on pci7 em0: Ethernet address: 00:04:23:c3:06:78
>> em0: [FAST]
>> skc0: <3Com 3C940 Gigabit Ethernet> port 0xe800-0xe8ff mem
>> 0xfebd8000-0xfebdbfff  irq 15 at device 6.0 on pci7
>> skc0: 3Com Gigabit NIC (3C2000) rev. (0x1)
>> sk0: <Marvell Semiconductor, Inc. Yukon> on skc0
>> sk0: Ethernet address: 00:0a:5e:65:ad:c3
>> miibus0: <MII bus> on sk0
>> e1000phy0: <Marvell 88E1000 Gigabit PHY> on miibus0
>> e1000phy0:  10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, 1000baseTX-FDX, auto
>>
>> P.S.: I am using the FreeBSD/amd64.
>>
>> Brian A. Seklecki wrote:
>>
>>> Show us netstat -m on the broken kernel?  Show us your dmesg(8) for em(4).
>>>
>>> TIA,
>>> ~BAS
>>>
>>> On Fri, 2007-03-30 at 11:13 -0300, Thiago Esteves de Oliveira wrote:
>>>
>>>> Hello,
>>>>
>>>> I've had a problem with one of my FreeBSD servers, the machine has stopped its network
services and then sent these messages:
>>>>
>>>> -Mar 27 13:00:03 anubis dhcpd: send_packet: No buffer space available -Mar 27 13:00:26 anubis
routed[431]: Send bcast sendto(em0,
>>>> 146.164.92.255.520): No buffer space available
>>>>
>>>> The messages were repeated a lot of times before a temporary solution. I've changed the
kernel(FreeBSD 6.2) to an older one(FreeBSD 6.1) and since then it's been working well. What
happened?
>>>>
>>>> P.S.: I can give more informations if necessary.
>>>>
>>>>





-- 
Thiago Esteves de Oliveira