netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Timo Teras <timo.teras@iki.fi>
To: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Ben Hutchings <bhutchings@solarflare.com>,
	netdev@vger.kernel.org, Francois Romieu <romieu@fr.zoreil.com>
Subject: Re: linux-3.0.18+r8169+ipv4/tcp forwarding = tso/gso weirdness and performance degration
Date: Wed, 14 Mar 2012 22:33:43 +0200	[thread overview]
Message-ID: <20120314223343.23dc9df3@vostro> (raw)
In-Reply-To: <1331755965.6022.55.camel@edumazet-glaptop>

On Wed, 14 Mar 2012 13:12:45 -0700 Eric Dumazet
<eric.dumazet@gmail.com> wrote:

> On Wed, 2012-03-14 at 21:51 +0200, Timo Teras wrote:
> 
> > Very good point. I thought gso also enabled gro, but seems that my
> > ethtool was old.
> > 
> > And GRO was enabled along with some other stuff. Turning GRO off
> > made my tcp performance immediately a lot better; jumped from 2MB/s
> > to 8MB/s or so (not ideal yet, though; but the remainder of the
> > difference could be related to other issue).
> > 
> > So something is definitely broke in 3.0.x with GRO enabled, but GSO
> > off.
> 
> "ifconfig eth2 ; netstat -s" can really help, I suspect tcp stack
> drops

After doing several wget's that have "bad performance".

# ifconfig eth2
eth2      Link encap:Ethernet  HWaddr 00:30:18:AB:6B:56  
          inet addr:10.26.0.2  Bcast:0.0.0.0  Mask:255.255.255.252
          inet6 addr: fe80::230:18ff:feab:6b56/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:32334060 errors:0 dropped:0 overruns:0 frame:0
          TX packets:18520452 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000 
          RX bytes:1775070027 (1.6 GiB)  TX bytes:1962364861 (1.8 GiB)
          Interrupt:16 Base address:0xc000 

# ethtool -S eth2
NIC statistics:
     tx_packets: 2069391193
     rx_packets: 3245815642
     tx_errors: 0
     rx_errors: 645238
     rx_missed: 31414
     align_errors: 0
     tx_single_collisions: 0
     tx_multi_collisions: 0
     unicast: 3245815640
     broadcast: 2
     multicast: 0
     tx_aborted: 0
     tx_underrun: 0

netstat is busybox, so no -s. But /proc/net/netstat as parsed looks
like:
SyncookiesSent 0 
SyncookiesRecv 0 
SyncookiesFailed 3 
EmbryonicRsts 0 
PruneCalled 0 
RcvPruned 0 
OfoPruned 0 
OutOfWindowIcmps 0 
LockDroppedIcmps 0 
ArpFilter 0 
TW 15 
TWRecycled 0 
TWKilled 0 
PAWSPassive 0 
PAWSActive 0 
PAWSEstab 0 
DelayedACKs 564 
DelayedACKLocked 5 
DelayedACKLost 0 
ListenOverflows 0 
ListenDrops 0 
TCPPrequeued 73 
TCPDirectCopyFromBacklog 605264 
TCPDirectCopyFromPrequeue 15961 
TCPPrequeueDropped 0 
TCPHPHits 191774 
TCPHPHitsToUser 425 
TCPPureAcks 19228 
TCPHPAcks 112359 
TCPRenoRecovery 0 
TCPSackRecovery 2 
TCPSACKReneging 0 
TCPFACKReorder 0 
TCPSACKReorder 4 
TCPRenoReorder 0 
TCPTSReorder 2 
TCPFullUndo 2 
TCPPartialUndo 107 
TCPDSACKUndo 0 
TCPLossUndo 0 
TCPLoss 0 
TCPLostRetransmit 0 
TCPRenoFailures 0 
TCPSackFailures 0 
TCPLossFailures 0 
TCPFastRetrans 2 
TCPForwardRetrans 2 
TCPSlowStartRetrans 0 
TCPTimeouts 0 
TCPRenoRecoveryFail 0 
TCPSackRecoveryFail 0 
TCPSchedulerFailed 0 
TCPRcvCollapsed 0 
TCPDSACKOldSent 0 
TCPDSACKOfoSent 0 
TCPDSACKRecv 4 
TCPDSACKOfoRecv 0 
TCPAbortOnSyn 0 
TCPAbortOnData 6 
TCPAbortOnClose 4 
TCPAbortOnMemory 0 
TCPAbortOnTimeout 0 
TCPAbortOnLinger 0 
TCPAbortFailed 0 
TCPMemoryPressures 0 
TCPSACKDiscard 0 
TCPDSACKIgnoredOld 0 
TCPDSACKIgnoredNoUndo 0 
TCPSpuriousRTOs 0 
TCPMD5NotFound 0 
TCPMD5Unexpected 0 
TCPSackShifted 0 
TCPSackMerged 0 
TCPSackShiftFallback 4484 
TCPBacklogDrop 0 
TCPMinTTLDrop 0 
TCPDeferAcceptDrop 0 
IPReversePathFilter 4 
TCPTimeWaitOverflow 0 

  reply	other threads:[~2012-03-14 20:34 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-03-14 17:01 linux-3.0.18+r8169+ipv4/tcp forwarding = tso/gso weirdness and performance degration Timo Teras
2012-03-14 17:15 ` Eric Dumazet
2012-03-14 17:29   ` Timo Teras
2012-03-14 18:25     ` Eric Dumazet
2012-03-14 19:29     ` Ben Hutchings
2012-03-14 19:51       ` Timo Teras
2012-03-14 20:12         ` Eric Dumazet
2012-03-14 20:33           ` Timo Teras [this message]
2012-03-14 20:52             ` Eric Dumazet
2012-03-14 20:53             ` Francois Romieu
2012-03-15  6:06               ` Timo Teras
2012-03-15 15:11                 ` Timo Teras
2012-03-15 16:11                   ` Eric Dumazet
2012-03-15 18:47                     ` Timo Teras
2012-03-15 19:11                   ` Francois Romieu
2012-03-16 20:15                     ` Timo Teras
2012-03-17  9:56                       ` Timo Teras
2012-03-17 11:35                         ` Francois Romieu
2012-03-17 22:20                           ` Francois Romieu
2012-03-18  7:00                             ` Timo Teras
2012-03-20 15:31                             ` Timo Teras
2012-03-20 18:20                               ` Francois Romieu
2012-03-14 21:16 ` Francois Romieu

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20120314223343.23dc9df3@vostro \
    --to=timo.teras@iki.fi \
    --cc=bhutchings@solarflare.com \
    --cc=eric.dumazet@gmail.com \
    --cc=netdev@vger.kernel.org \
    --cc=romieu@fr.zoreil.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).