Re: ix(intel) vs mlxen(mellanox) 10Gb performance

> On Aug 18, 2015, at 12:49 AM, Rick Macklem wrote: > > Daniel Braniss wrote: >> >>> On Aug 17, 2015, at 3:21 PM, Rick Macklem wrote: >>> >>> Daniel Braniss wrote: > On Aug 17, 2015, at 1:41 PM, Christopher Forgeron > wrote: > > FYI, I can regularly hit 9.3 Gib/s with

pf and new interface

I have the following rule in pf.conf: set skip on tap and even the following one: set skip on tap0 The rules are loaded at the system start-up time, but the tap interface may not be created until much later. When tap0 is first created the skip rules are not applied to it and the traffic gets fil

Re: pf and new interface

  --- Original message --- From: "Andriy Gapon" Date: 18 August 2015, 14:05:15 > I have the following rule in pf.conf: > set skip on tap > and even the following one: > set skip on tap0 > > The rules are loaded at the system start-up time, but the tap interface > may not be created until m

Re: pf and new interface

On 18/08/2015 14:18, wishmaster wrote: > --- Original message --- > From: "Andriy Gapon" > Date: 18 August 2015, 14:05:15 > > >> I have the following rule in pf.conf: >> set skip on tap >> and even the following one: >> set skip on tap0 >> >> The rules are loaded at the system start-up time, b

Re[2]: pf and new interface

--- Original message --- From: "Andriy Gapon" Date: 18 August 2015, 14:35:36 > On 18/08/2015 14:18, wishmaster wrote: > > --- Original message --- > > From: "Andriy Gapon" > > Date: 18 August 2015, 14:05:15 > > > > > >> I have the following rule in pf.conf: > >> set skip on tap > >>

Re: pf and new interface

On 18/08/2015 14:55, wishmaster wrote: > --- Original message --- > From: "Andriy Gapon" > Date: 18 August 2015, 14:35:36 > > > >> On 18/08/2015 14:18, wishmaster wrote: >>> --- Original message --- >>> From: "Andriy Gapon" >>> Date: 18 August 2015, 14:05:15 >>> >>> I have the follow

Re: ix(intel) vs mlxen(mellanox) 10Gb performance

Daniel Braniss wrote: > > > On Aug 18, 2015, at 12:49 AM, Rick Macklem wrote: > > > > Daniel Braniss wrote: > >> > >>> On Aug 17, 2015, at 3:21 PM, Rick Macklem wrote: > >>> > >>> Daniel Braniss wrote: > > > On Aug 17, 2015, at 1:41 PM, Christopher Forgeron > > > > wrote: >

Re: ix(intel) vs mlxen(mellanox) 10Gb performance

On 08/18/15 14:53, Rick Macklem wrote: If this is just a test machine, maybe you could test with these lines (at about #880) in sys/netinet/tcp_output.c commented out? (It looks to me like this will disable TSO for almost all the NFS writes.) - around line #880 in sys/netinet/tcp_output.c:

Re: ix(intel) vs mlxen(mellanox) 10Gb performance

On 08/18/15 14:53, Rick Macklem wrote: 2572 ifp->if_hw_tsomax = 65518; >2573 ifp->if_hw_tsomaxsegcount = IXGBE_82599_SCATTER; >2574 ifp->if_hw_tsomaxsegsize = 2048; Hi, If IXGBE_82599_SCATTER is the maximum scatter/gather ent

Re: ix(intel) vs mlxen(mellanox) 10Gb performance

sorry, it’s been a tough day, we had a major meltdown, caused by a faulty gbic :-( anyways, could you tell me what to do? comment out, fix the off by one? the machine is not yet production. thanks, danny > On 18 Aug 2015, at 16:32, Hans Petter Selasky wrote: > > On 08/18/15 14:53, Ric

Re: ix(intel) vs mlxen(mellanox) 10Gb performance

On Tue, Aug 18, 2015 at 05:09:41PM +0300, Daniel Braniss wrote: > sorry, it's been a tough day, we had a major meltdown, caused by a faulty > gbic :-( > anyways, could you tell me what to do? > comment out, fix the off by one? > > the machine is not yet production. Can you collect this informat

Re: Poor high-PPS performance of the 10G ixgbe(9) NIC/driver in FreeBSD 10.1

Yes, we've confirmed it's IXGBE_FDIR. That's good it comes disabled in 10.2. Thanks everyone for constructive input! -Max ___ freebsd-net@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-net To unsubscribe, send any mail to "f

Re: Poor high-PPS performance of the 10G ixgbe(9) NIC/driver in FreeBSD 10.1

you're welcome. Someone should really add a release errata to 10.1 or something. -a On 18 August 2015 at 10:59, Maxim Sobolev wrote: > Yes, we've confirmed it's IXGBE_FDIR. That's good it comes disabled in 10.2. > > Thanks everyone for constructive input! > > -Max > __

Re: pf and new interface

Hmm does the: set skip on (tap) syntax work in this case? Basically parentheses around the alias should tell pf that the IP is volatile and can be either activated at later time or it can be dynamic via dhcp etc. -Reko ___ freebsd-net@freebsd.or

Re: Poor high-PPS performance of the 10G ixgbe(9) NIC/driver in FreeBSD 10.1

On Tue, Aug 18, 2015 at 11:18:33AM -0700, hiren panchasara wrote: > On 08/18/15 at 11:03P, Adrian Chadd wrote: > > you're welcome. > > > > Someone should really add a release errata to 10.1 or something. > > Yes, I strongly feel the same. Adding gjb@ here to see how that can be > done. > Please

Re: pf and new interface

On 18/08/2015 20:43, Reko Turja wrote: > Hmm does the: > > set skip on (tap) > > syntax work in this case? Basically parentheses around the alias should > tell pf that the IP is volatile and can be either activated at later > time or it can be dynamic via dhcp etc. I will check and follow up. -

Re: Poor high-PPS performance of the 10G ixgbe(9) NIC/driver in FreeBSD 10.1

On 08/18/15 at 06:25P, Glen Barber wrote: > On Tue, Aug 18, 2015 at 11:18:33AM -0700, hiren panchasara wrote: > > On 08/18/15 at 11:03P, Adrian Chadd wrote: > > > you're welcome. > > > > > > Someone should really add a release errata to 10.1 or something. > > > > Yes, I strongly feel the same. Ad

Re: Poor high-PPS performance of the 10G ixgbe(9) NIC/driver in FreeBSD 10.1

On 08/18/15 at 11:03P, Adrian Chadd wrote: > you're welcome. > > Someone should really add a release errata to 10.1 or something. Yes, I strongly feel the same. Adding gjb@ here to see how that can be done. Cheers, Hiren > > > -a > > > On 18 August 2015 at 10:59, Maxim Sobolev wrote: > > Ye

Re: ix(intel) vs mlxen(mellanox) 10Gb performance

Hans Petter Selasky wrote: > On 08/18/15 14:53, Rick Macklem wrote: > > 2572 ifp->if_hw_tsomax = 65518; > >> >2573 ifp->if_hw_tsomaxsegcount = > >> >IXGBE_82599_SCATTER; > >> >2574 ifp->if_hw_tsomaxsegsize = 2048; > > Hi, > > If

Re: ix(intel) vs mlxen(mellanox) 10Gb performance

Hans Petter Selasky wrote: > On 08/18/15 14:53, Rick Macklem wrote: > > If this is just a test machine, maybe you could test with these lines (at > > about #880) > > in sys/netinet/tcp_output.c commented out? (It looks to me like this will > > disable TSO > > for almost all the NFS writes.) > > - a

Panic [page fault] in _ieee80211_crypto_delkey(): stable/10/amd64 @r286878

I was minding my own business in a staff meeting this afternoon, and my laptop rebooted; seems it got a panic. I've copied the core.txt.0 file to , along with a verbose dmesg.boot from this morning and output of "pciconf -l -v". This was runnin

Re: ix(intel) vs mlxen(mellanox) 10Gb performance

Daniel Braniss wrote: > > > On Aug 18, 2015, at 12:49 AM, Rick Macklem wrote: > > > > Daniel Braniss wrote: > >> > >>> On Aug 17, 2015, at 3:21 PM, Rick Macklem wrote: > >>> > >>> Daniel Braniss wrote: > > > On Aug 17, 2015, at 1:41 PM, Christopher Forgeron > > > > wrote: >

Re: ix(intel) vs mlxen(mellanox) 10Gb performance

On 08/18/15 23:54, Rick Macklem wrote: Ouch! Yes, I now see that the code that counts the # of mbufs is before the code that adds the tcp/ip header mbuf. In my opinion, this should be fixed by setting if_hw_tsomaxsegcount to whatever the driver provides - 1. It is not the driver's responsibility