Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 25 Jan 2013 17:05:01 +0100
From:      Fleuriot Damien <ml@my.gd>
To:        Daniel Braniss <danny@cs.huji.ac.il>
Cc:        "freebsd-stable@freebsd.org" <freebsd-stable@freebsd.org>
Subject:   Re: bge bad performance
Message-ID:  <AB4851B1-63E0-4A95-9083-6D184B6ADFD9@my.gd>
In-Reply-To: <E1Tyle4-000DG8-NZ@kabab.cs.huji.ac.il>
References:  <E1Tygmo-0009Vr-7W@kabab.cs.huji.ac.il> <727F9491-A88F-4A4E-BD4F-61235F31500E@my.gd> <E1Tyle4-000DG8-NZ@kabab.cs.huji.ac.il>

next in thread | previous in thread | raw e-mail | index | archive | help

On Jan 25, 2013, at 4:56 PM, Daniel Braniss <danny@cs.huji.ac.il> wrote:

>>> On 25 Jan 2013, at 11:45, Daniel Braniss <danny@cs.huji.ac.il> =
wrote:
>>=20
>>> Hi,
>>> It seems that I have more issues with the bge,
>>>    Broadcom NetXtreme Gigabit Ethernet, ASIC rev. 0x5720000
>>>=20
>>> ifconfig says:
>>> bge2: flags=3D8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> metric 0 =
mtu 15=3D00
>>> =
options=3Dc019b<RXCSUM,TXCSUM,VLAN_MTU,VLAN_HWTAGGING,VLAN_HWCSUM,TSO4,VLA=
=3DN_HWTS
>>> O,LINKSTATE>
>>> ether xx...
>>> inet6 xxx prefixlen 64 scopeid 0x3
>>> inet xxx... netmask 0xfffff000 broadcast yyy
>>> nd6 options=3D21<PERFORMNUD,AUTO_LINKLOCAL>
>>> media: Ethernet autoselect (1000baseT <full-duplex>)
>>> status: active
>>>=20
>>> iperf reports
>>> [  3]  0.0-10.0 sec   230 MBytes   193 Mbits/sec
>>>=20
>>=20
>> Post your iperf command line and explain the test environment, as in =
direct connection vs switched network link ?
>>=20
>=20
> I run iperf -s on the same host for a long time, so not to
> influence the results.
> the network is switched, and I try out several other host to be able =
to
> a- compare
> b- detect if the network is too busy.
> c- to somehow 'normalize'
>=20
> what promped me to check iperf in this case was that tar via NFS =
started to=20
> give
> read timeouts.
>=20
> so I run perf -c:
> with=20
> dev.bge.2.forced_collapse: 0 or 1
> dev.bge.2.msi: 1
> pe-04> iperf -c minbari
> ------------------------------------------------------------
> Client connecting to minbari, TCP port 5001
> TCP window size:  257 KByte (default)
> ------------------------------------------------------------
> [  3] local 132.65.16.35 port 12835 connected with 132.65.16.212 port =
5001
> [ ID] Interval       Transfer     Bandwidth
> [  3]  0.0-10.0 sec   238 MBytes   200 Mbits/sec
>=20
> setting:
> dev.bge.2.forced_collapse:1
> dev.bge.2.msi:1
> pe-04> iperf -c minbari
> ------------------------------------------------------------
> Client connecting to minbari, TCP port 5001
> TCP window size:  257 KByte (default)
> ------------------------------------------------------------
> [  3] local 132.65.16.35 port 23207 connected with 132.65.16.212 port =
5001
> [ ID] Interval       Transfer     Bandwidth
> [  3]  0.0-10.0 sec   506 MBytes   425 Mbits/sec
>=20
> the same command from a similar host, but with bce:
> Client connecting to minbari, TCP port 5001
> TCP window size:  257 KByte (default)
> ------------------------------------------------------------
> [  3] local 132.65.80.2 port 30664 connected with 132.65.16.212 port =
5001
> [ ID] Interval       Transfer     Bandwidth
> [  3]  0.0-10.0 sec   976 MBytes   818 Mbits/sec
>=20
> In case it's not obvious, all hosts are server class, the network =
shows no=20
> errors,
> neither the hosts.
>=20
> cheers
> 	danny


Just did a quick roundup of our boxes and all I've got is bce, re and em =
interfaces, I won't be able to help you much :(

Try playing with iperf's args, like the buffers and number of // =
connections ?




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?AB4851B1-63E0-4A95-9083-6D184B6ADFD9>