ПРОЕКТЫ 

  АРХИВ 

Apache-Talk @lexa.ru 

Inet-Admins @info.east.ru 

Filmscanners @halftone.co.uk 

Security-alerts @yandex-team.ru 

nginx-ru @sysoev.ru 

  СТАТЬИ 

  ПЕРСОНАЛЬНОЕ 

  ПРОГРАММЫ 

ПИШИТЕ
ПИСЬМА












     АРХИВ :: nginx-ru
Nginx-ru mailing list archive (nginx-ru@sysoev.ru)
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [SPAM]Re: уменьшение тр афика в сети

забыл сказать
после загрузки сервера сделал обоим em-интерфейсам ifconfig emX polling

On 19.03.2009, at 18:46, Alexey V. Karagodov wrote:


# w
6:42PM  up 3 days, 13:49, 3 users, load averages: 0.06, 0.10, 0.13
em0 9000 <Link#1> 00:30:48:74:51:16 153436664 0 206056891 0 0 em1 9000 <Link#2> 00:30:48:74:51:16 54944561 0 99091199 0 0 lagg0 9000 <Link#6> 00:30:48:74:51:16 208360580 0 305126939 3870 0 

на всех интерфейсах ошибок 0

настройки:

# cat /etc/sysctl.conf

# $FreeBSD: src/etc/sysctl.conf,v 1.8 2003/03/13 18:43:50 mux Exp $
#
# This file is read when going to multi-user and its contents piped thru # ``sysctl'' to adjust kernel values. ``man 5 sysctl.conf'' for details. 
#
# Uncomment this to prevent users from seeing information about processes that 
# are being run under another UID.
#security.bsd.see_other_uids=0
#net.link.ether.inet.log_arp_wrong_iface=0

dev.em.0.rx_processing_limit=-1
dev.em.0.rx_int_delay=0
dev.em.0.tx_int_delay=0
dev.em.0.rx_abs_int_delay=0
dev.em.0.tx_abs_int_delay=0
dev.em.1.rx_processing_limit=-1
dev.em.1.rx_int_delay=0
dev.em.1.tx_int_delay=0
dev.em.1.rx_abs_int_delay=0
dev.em.1.tx_abs_int_delay=0

net.inet.ip.intr_queue_maxlen=4096
net.isr.direct=1
net.inet.tcp.tso=1

kern.polling.burst_max=1000

kern.threads.max_threads_per_proc=8192

kern.ipc.maxsockets=204800
kern.ipc.somaxconn=4096
#kern.maxfiles=204800
#kern.maxfilesperproc=200000
net.inet.ip.portrange.first=1024
net.inet.ip.portrange.last=65535
net.inet.ip.portrange.randomized=0
net.inet.tcp.blackhole=1
net.inet.tcp.fast_finwait2_recycle=1
net.inet.tcp.maxtcptw=40960
net.inet.tcp.msl=30000
net.inet.tcp.nolocaltimewait=1
net.inet.tcp.recvspace=131072
net.inet.tcp.sendspace=131072
net.inet.tcp.syncookies=1
net.inet.udp.blackhole=1
net.inet.udp.recvspace=131072

#kern.ipc.nmbclusters=262144
#kern.ipc.nmbclusters=0
#net.inet.tcp.recvspace=8192
#net.inet.tcp.sendspace=16384

machdep.hyperthreading_allowed=1
#vfs.nfs.access_cache_timeout=5
#We did some tuning already and our current sysctl.conf looks like this: 

#kern.ipc.somaxconn=32768
#net.inet.icmp.icmplim=3000
#kern.ipc.maxsockets=300000
#net.inet.tcp.delayed_ack=1
#net.inet.tcp.finwait2_timeout=15000
#net.inet.tcp.maxtcptw=196607
#dev.em.0.rx_processing_limit=-1
#net.isr.direct=0

#hw.acpi.cpu.cx_lowest=C3
# reduce swap paging
#vm.defer_swapspace_pageouts=1
# other tuning from "man tuning"
#kern.ipc.shm_use_phys=1

#

# cat /boot/loader.conf

verbose_loading="YES"
loader_logo="beastie"
#ichsmb_load="YES"
#smb_load="YES"
#ichwd_load="YES"
ng_ether_load="YES"
#linux_load="YES"
accf_data_load="YES"
accf_http_load="YES"

net.inet.tcp.syncache.hashsize=1024
net.inet.tcp.syncache.bucketlimit=100
net.inet.tcp.tcbhashsize=4096
kern.ipc.nsfbufs=10240
kern.ipc.nmbclusters=0
vm.kmem_size=1G
vm.kmem_size_max=1G

hw.em.tx_int_delay="0"
hw.em.rx_int_delay="0"
hw.em.tx_abs_int_delay="0"
hw.em.rx_abs_int_delay="0"
hw.em.rxd="4096"
hw.em.txd="4096"
#hw.em.smart_pwr_down=""
#hw.em.sbp=""
hw.em.enable_msi="1"
hw.em.rx_process_limit="-1"

#

# cat /etc/rc.conf
#etc
linux_enable="YES"
#watchdogd="YES"
blanktime="60"
check_quotas="NO"
keyrate="fast"
moused_enable="YES"
named_enable="YES"
named_flags="-c /etc/namedb/include/named.conf"
dhcpd_enable="YES"
dhcpd_flags="-q"
dhcpd_conf="/usr/local/etc/dhcpd.conf"
dhcpd_ifaces=""
dhcpd_withumask="022"
nfs_reserved_port_only="YES"
rpcbind_enable="YES"
saver="green"
sshd_enable="YES"
usbd_enable="YES"
ntpd_enable="YES"
ntpd_sync_on_start="YES"
gateway_enable="YES"
#router_enable="YES"
router_flags="-v -s"
defaultrouter="x.x.x.x"
hostname="frontend1.x.x.x"
#ipcad_enable="YES"
#apcupsd_enable="YES"
nginx_enable="YES"
#squid_enable="YES"
sendmail_enable="NO"
sendmail_submit_enable="NO"
sendmail_outbound_enable="NO"
sendmail_msp_queue_enable="NO"
#inetd_enable="YES"
zabbix_agentd_enable="YES"

#locale
mousechar_start="3"
#keymap="ru.cp1251"
#scrnmap="win2cpp866"
#font8x16="cp866-8x16"
#font8x14="cp866-8x14"
#font8x8="cp866-8x8"

ifconfig_em0="up mtu 9000 vlanmtu vlanhwtag txcsum rxcsum tso lro wol"
ifconfig_em1="up mtu 9000 vlanmtu vlanhwtag txcsum rxcsum tso lro wol"

cloned_interfaces="lagg0 ..."

ifconfig_lagg0="laggproto lacp laggport em0 laggport em1"
ifconfig_lagg0_alias0=...


On 19.03.2009, at 16:57, Монашёв Михаил wrote:


Здравствуйте, Дмитрий.

??>> А там можно какие-нить sysctl-ки покрутить?


/boot/loader.conf:



hw.em.rxd="4096"
hw.em.txd="4096"
hw.em.rx_process_limit="-1"


ещё добавил net.inet.ip.intr_queue_maxlen=4096 в /etc/sysctl.conf
На двух машинах Ierrs в netstat -i пропали. А на одной, с самым большим количеством пакетов, остались. Растут правда намного медленнее, но растут. за 5 часов на 3500 где-то. Это конечно не много, 
но всёравно не хорошо.

Картина вот такая сейчас:

netstat -i -Iem1
Name Mtu Network Address Ipkts Ierrs Opkts Oerrs Coll em1 9216 <Link#2> 00:04:23:de:b0:1c 281537845 3586 147376712 0 0 
а за какое время данные?




Mar 19 16:49:45 softsearch1 kernel: em1: Excessive collisions = 0
Mar 19 16:49:45 softsearch1 kernel: em1: Sequence errors = 0
Mar 19 16:49:45 softsearch1 kernel: em1: Defer count = 0
Mar 19 16:49:45 softsearch1 kernel: em1: Missed Packets = 3586
Mar 19 16:49:45 softsearch1 kernel: em1: Receive No Buffers = 1428
Mar 19 16:49:45 softsearch1 kernel: em1: Receive Length Errors = 0
Mar 19 16:49:45 softsearch1 kernel: em1: Receive errors = 0
Mar 19 16:49:45 softsearch1 kernel: em1: Crc errors = 0
Mar 19 16:49:45 softsearch1 kernel: em1: Alignment errors = 0
Mar 19 16:49:45 softsearch1 kernel: em1: Collision/Carrier extension errors = 0 
Mar 19 16:49:45 softsearch1 kernel: em1: RX overruns = 9
Mar 19 16:49:45 softsearch1 kernel: em1: watchdog timeouts = 0
Mar 19 16:49:45 softsearch1 kernel: em1: RX MSIX IRQ = 0 TX MSIX IRQ = 0 LINK MSIX IRQ = 0 
Mar 19 16:49:45 softsearch1 kernel: em1: XON Rcvd = 0
Mar 19 16:49:45 softsearch1 kernel: em1: XON Xmtd = 0
Mar 19 16:49:45 softsearch1 kernel: em1: XOFF Rcvd = 0
Mar 19 16:49:45 softsearch1 kernel: em1: XOFF Xmtd = 0
Mar 19 16:49:45 softsearch1 kernel: em1: Good Packets Rcvd = 285783391 Mar 19 16:49:45 softsearch1 kernel: em1: Good Packets Xmtd = 149632906 
Mar 19 16:49:45 softsearch1 kernel: em1: TSO Contexts Xmtd = 0
Mar 19 16:49:45 softsearch1 kernel: em1: TSO Contexts Failed = 0

Это данные примерно за 5 часов.

Может имеет смысл уменьшить dev.em.1.rx_abs_int_delay? Сейчас оно 66.

--
С уважением,
Монашёв Михаил, SoftSearch.ru
mailto:postmaster@xxxxxxxxxxxxx
ICQ# 166233339
http://michael.mindmix.ru/
Без бэкапа по жизни.

 



Copyright © Lexa Software, 1996-2009.