full socket buffers
Mirek Lauš
mirek.laus at gmail.com
Sat Dec 18 20:25:02 UTC 2010
On Sat, Dec 18, 2010 at 9:08 PM, <sthaug at nethelp.no> wrote:
>> we're running ISC DHCP 4.1.1p1 on FreeBSD 7.3-RELEASE box
>> for about 30k hosts with various leasetime from 4 hours to 7 days.
>> Time to time we experience significant performance loss while
>> clients are unable to get a lease. In the time the UDP full socket
>> buffer is constantly growing:
> ...
>> How could I tell the cause of the full buffer socket?
>
> How is your disk utilization? Good ISC DHCP performance is critically
> dependent on good disk performance.
>
> We're also running 4.1.1p1, on FreeBSD 7.3-STABLE. Server hardware is
> Dell PE 1850, with two SCSI disks in a RAID-1 configuration. We have
> around 100k leases on two servers in a failover pair, with most of the
> pools on 24 hour leasetime. We don't see the problem you're seeing.
>
> Steinar Haug, Nethelp consulting, sthaug at nethelp.no
>
We're running the dhcpd service on two IBM x3550 servers in failover,
both have two 74GB 15k disks in RAID-1 so very similar to yours.
iostat sample:
[root at dhcp1 ~]# iostat 1
tty aacd0 cpu
tin tout KB/t tps MB/s us ni sy in id
0 131 16.80 20 0.34 -20 -0 -1 -0 121
0 242 16.00 19 0.30 1 0 0 0 99
0 392 16.00 27 0.42 1 0 1 0 98
0 155 16.00 18 0.28 1 0 0 0 98
0 152 16.00 16 0.25 2 0 0 0 98
0 153 16.00 23 0.36 0 0 0 0 99
0 262 16.00 9 0.14 1 0 0 0 99
0 43 16.00 16 0.25 1 0 0 0 99
0 43 10.26 39 0.39 0 0 1 0 99
0 263 16.00 22 0.34 1 0 0 0 99
0 287 16.00 5 0.08 1 0 0 0 99
0 43 16.00 29 0.45 1 0 0 0 99
0 151 16.00 77 1.20 1 0 0 0 99
0 43 16.00 26 0.41 1 0 0 0 98
0 43 16.00 22 0.34 1 0 0 0 99
0 43 16.00 11 0.17 0 0 0 0 100
0 44 16.00 22 0.34 1 0 0 0 99
0 154 16.00 29 0.45 2 0 1 0 98
0 166 16.00 13 0.20 1 0 0 0 99
0 209 16.00 18 0.28 1 0 0 0 99
0 43 16.00 7 0.11 1 0 0 0 99
-m
More information about the dhcp-users
mailing list