From mboxrd@z Thu Jan 1 00:00:00 1970 From: "M. Istehbab" Subject: TCP connection failure in kernel 2.6.25 Date: Sat, 26 Apr 2008 14:17:22 +0500 Message-ID: Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit To: netdev@vger.kernel.org Return-path: Received: from wf-out-1314.google.com ([209.85.200.174]:19555 "EHLO wf-out-1314.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750866AbYDZJRW (ORCPT ); Sat, 26 Apr 2008 05:17:22 -0400 Received: by wf-out-1314.google.com with SMTP id 28so3222053wff.4 for ; Sat, 26 Apr 2008 02:17:22 -0700 (PDT) Content-Disposition: inline Sender: netdev-owner@vger.kernel.org List-ID: Package: linux-image-2.6.24-1-686 Version: 2.6.24-6 Severity: critical I have tried it with kernels from debian between 2.6.18-5-686 - 2.6.25. Be it vanilla versions/sources from kernel.org or debian built kernels. I have the same issue. The problem is random TCP connection failure for ipv4. I have seen a severe issue of this problem in my transparent proxy setup. This problem also occurs at times when it is not redirecting traffic, rather accepting it directly. Increasing the number of file descriptors didn't help. I rebuilt the kernel with 65K file descriptors, still the problem persists. Even if it is one client behind squid or this box, this happens. It will at random, not accept connections. squid doesn't report any in the log. netstat doesn't show any in its output as well. I have changed NICs, I have changed between squid versions from 2.x to 3.x. Nothing has helped. Following are some settings to mention; ulimit -n is set to 8000 net.core.somaxconn = 1024 net.core.netdev_max_backlog = 3000 net.ipv4.tcp_max_syn_backlog = 3000 squid is built with 8000 file descriptors. This is a 60 client LAN behind this box, using it as an internet gateway/firewall for the following traffic; 1) vpn 2) email 3) cvs 4) browsing (http/https) 5) some ftp ///////////////////////////////////////////////////////////////////////////////////////// This is a PIII (Coppermine) cpu MHz : 864.499 RAM : 1 GB Hard disk : 2x 36 GB SCSI hard drives RAID : raid1 Non Raid part : 3 non raid partitions ///////////////////////////////////////////////////////////////////////////////////////// # cat /proc/mst cat: /proc/mst: No such file or directory sglhrcvs:~/work/traffic-shaping# cat /proc/mdstat Personalities : [raid1] md0 : active raid1 sda1[0] sdb1[1] 32226240 blocks [2/2] [UU] unused devices: ///////////////////////////////////////////////////////////////////////////////////////// This box worked fine when it had RHEL WS3 on it. -- Package-specific info: ** Version: Linux version 2.6.24-1-686 (Debian 2.6.24-6) (waldi@debian.org) (gcc version 4.1.3 20080308 (prerelease) (Debian 4.1.2-21)) #1 SMP Sat Apr 19 00:37:55 UTC 2008 ** Command line: root=/dev/md0 ro process `sysctl' is using deprecated sysctl (syscall) net.ipv6.neigh.default.retrans_time; Use net.ipv6.neigh.default.retrans_time_ms instead. ** Loaded modules: Module Size Used by cls_u32 8100 1 sch_cbq 16352 1 tcp_diag 1792 0 inet_diag 12136 1 tcp_diag xt_limit 2656 1 xt_state 2464 1 ipt_REDIRECT 2080 3 ipt_MASQUERADE 3776 3 xt_MARK 2272 11 iptable_nat 6916 1 nf_nat 18316 3 ipt_REDIRECT,ipt_MASQUERADE,iptable_nat nf_conntrack_ipv4 17352 3 iptable_nat nf_conntrack 62272 5 xt_state,ipt_MASQUERADE,iptable_nat,nf_nat,nf_conntrack_ipv4 iptable_filter 2976 1 ppdev 8804 0 lp 11076 0 ac 6116 0 battery 13572 0 xt_TCPMSS 4384 1 xt_tcpmss 2240 1 xt_tcpudp 3136 29 iptable_mangle 2784 1 ip_tables 13188 3 iptable_nat,iptable_filter,iptable_mangle x_tables 14244 10 xt_limit,xt_state,ipt_REDIRECT,ipt_MASQUERADE,xt_MARK,iptable_nat,xt_TCPMSS,xt_tcpmss,xt_tcpudp,ip_tables pppoe 12864 0 pppox 3980 1 pppoe ppp_generic 26196 2 pppoe,pppox slhc 5920 1 ppp_generic ipv6 240836 35 dm_snapshot 16964 0 dm_mirror 21600 0 dm_mod 55812 2 dm_snapshot,dm_mirror loop 16804 0 parport_pc 33668 1 parport 34280 3 ppdev,lp,parport_pc floppy 54628 0 rtc 13052 0 psmouse 36464 0 pcspkr 3200 0 i2c_i801 9232 0 i2c_core 22432 1 i2c_i801 button 8432 0 serio_raw 6660 0 shpchp 31028 0 intel_agp 23412 1 agpgart 31688 1 intel_agp pci_hotplug 27712 1 shpchp intel_rng 5632 0 iTCO_wdt 11268 0 evdev 11104 1 sg 33264 0 sr_mod 16708 0 cdrom 32512 1 sr_mod ext3 122888 3 jbd 43732 1 ext3 mbcache 8288 1 ext3 raid1 22240 1 md_mod 73716 2 raid1 ide_generic 1280 0 [permanent] sd_mod 27200 7 aic7xxx 157560 5 scsi_transport_spi 22816 1 aic7xxx ata_generic 7428 0 libata 144464 1 ata_generic scsi_mod 141164 6 sg,sr_mod,sd_mod,aic7xxx,scsi_transport_spi,libata generic 4388 0 [permanent] 3c59x 41064 0 e100 33836 0 mii 5312 2 3c59x,e100 piix 7492 0 [permanent] uhci_hcd 23376 0 ide_core 108292 3 ide_generic,generic,piix usbcore 132940 2 uhci_hcd thermal 16028 0 processor 36520 1 thermal fan 4772 0 ** PCI devices: 00:00.0 Host bridge [0600]: Intel Corporation 82840 840 (Carmel) Chipset Host Bridge (Hub A) [8086:1a21] (rev 01) Subsystem: Dell Unknown device [1028:0096] Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- SERR- 00:01.0 PCI bridge [0604]: Intel Corporation 82840 840 (Carmel) Chipset AGP Bridge [8086:1a23] (rev 01) (prog-if 00 [Normal decode]) Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- Status: Cap- 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- SERR- TAbort- Reset- FastB2B- 00:1e.0 PCI bridge [0604]: Intel Corporation 82801AA PCI Bridge [8086:2418] (rev 02) (prog-if 00 [Normal decode]) Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- Status: Cap- 66MHz- UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- SERR- TAbort- Reset- FastB2B- 00:1f.0 ISA bridge [0601]: Intel Corporation 82801AA ISA Bridge (LPC) [8086:2410] (rev 02) Control: I/O+ Mem+ BusMaster+ SpecCycle+ MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- Status: Cap- 66MHz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- SERR- TAbort- SERR- TAbort- SERR- TAbort- SERR- TAbort- SERR- TAbort- SERR- TAbort- SERR- TAbort- SERR- TAbort- SERR- TAbort- SERR- TAbort- SERR- TAbort- Reset- FastB2B- Capabilities: [dc] Power Management version 1 Flags: PMEClk- DSI- D1- D2- AuxCurrent=220mA PME(D0-,D1-,D2-,D3hot-,D3cold-) Status: D0 PME-Enable- DSel=0 DScale=0 PME- Bridge: PM- B3+ 03:0a.0 SCSI storage controller [0100]: Adaptec AIC-7899P U160/m [9005:00cf] (rev 01) Subsystem: Dell Unknown device [1028:0096] Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV+ VGASnoop- ParErr- Stepping- SERR+ FastB2B- Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- SERR- TAbort- SERR-