ALT Linux sysadmins discussion
 help / color / mirror / Atom feed
From: ahtoh@pulnet.ru
To: ALT Linux sysadmin discuss <sysadmins@lists.altlinux.org>
Subject: Re: [Sysadmins] ksoftirq 99% cpu load.
Date: Tue, 14 Oct 2008 20:21:28 +0400
Message-ID: <200810142021.28231.ahtoh@pulnet.ru> (raw)
In-Reply-To: <48F4BFFE.7030700@unsafe.ru>

В сообщении от 14 октября 2008 Konstantin A. Lepikhov написал(a):
> 14.10.2008 18:13, ahtoh@pulnet.ru пишет:
> > irqbalance может распределяет прерывания например
> > от eth0 на CPU0, а от eth1 на CPU1 итд.
> > А здесь сразу по всем ядрам, что конечно лучше.
> > _______________________________________________
> > Sysadmins mailing list
> > Sysadmins@lists.altlinux.org
> > https://lists.altlinux.org/mailman/listinfo/sysadmins
>
> Вы это сами только что придумали?
> _______________________________________________
> Sysadmins mailing list
> Sysadmins@lists.altlinux.org
> https://lists.altlinux.org/mailman/listinfo/sysadmins

Вот /proc/interrupts c запущенным irqbalance ядро 2.6.18

           CPU0       CPU1       CPU2       CPU3
  0:   14736893          0     253682    4383039    IO-APIC-edge  timer
  1:       2393          0          0          0    IO-APIC-edge  i8042
  6:          3          0          0          0    IO-APIC-edge  floppy
  8:          1          0          0          0    IO-APIC-edge  rtc
  9:          1          0          0          0   IO-APIC-level  acpi
 14:         61          0          0          0    IO-APIC-edge  ide0
 58:  339791237  110911999          0          0         PCI-MSI  eth0
 66:  437172012          0          0          0         PCI-MSI  eth1
169:     175988          0       1648    2087386   IO-APIC-level  ioc0
NMI:      35854       9985        453        541
LOC:   19374644   19375064   19375160   19374988
ERR:          0
MIS:          0

Как видно прерывания от определенных системных устройств
распределены строго по определённым ядрам.

А вот 2.6.24-std-smp-def

           CPU0       CPU1       CPU2       CPU3
  0:         29          0          0          0   IO-APIC-edge      timer
  1:          1          0          1          0   IO-APIC-edge      i8042
  8:          0          1          0          0   IO-APIC-edge      rtc
  9:          0          0          0          1   IO-APIC-fasteoi   acpi
 12:          0          1          1          2   IO-APIC-edge      i8042
 17:          0          0          0          0   IO-APIC-fasteoi   
uhci_hcd:usb3, ehci_hcd:usb4
 18:          0          0          0          0   IO-APIC-fasteoi   libata, 
uhci_hcd:usb1, uhci_hcd:usb7
 19:          0          0          0          0   IO-APIC-fasteoi   
uhci_hcd:usb6
 21:      88026      87808      88783      89463   IO-APIC-fasteoi   libata, 
libata, uhci_hcd:usb2
 23:          0          0          0          0   IO-APIC-fasteoi   
uhci_hcd:usb5, ehci_hcd:usb8
2298: 1602881384 1602881600 1602880626 1602879945   PCI-MSI-edge      eth0
NMI:          0          0          0          0   Non-maskable interrupts
LOC:   35594469   15987382   18652239    9478400   Local timer interrupts
RES:        666       1473        745       1531   Rescheduling interrupts
CAL:       1425       1446       1447        191   function call interrupts
TLB:       1421       1537       1471       1578   TLB shootdowns
TRM:          0          0          0          0   Thermal event interrupts
THR:          0          0          0          0   Threshold APIC interrupts
SPU:          0          0          0          0   Spurious interrupts
ERR:          0

Видно, что все ядра поочередно учавствуют в обработке прерываний от одного 
устройства.

Или я не прав?


  reply	other threads:[~2008-10-14 16:21 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-10-14  9:48 ahtoh
2008-10-14 10:00 ` Led
2008-10-14 10:28 ` Max Ivanov
2008-10-14 11:06 ` Alex
2008-10-14 12:57 ` ahtoh
2008-10-14 13:35   ` Unsafe.ru
2008-10-14 13:45     ` ahtoh
2008-10-14 13:51       ` Konstantin A. Lepikhov
2008-10-14 14:13         ` ahtoh
2008-10-14 15:51           ` Konstantin A. Lepikhov
2008-10-14 16:21             ` ahtoh [this message]
2008-10-14 18:59               ` Led
2008-10-15  4:53                 ` ahtoh
2008-10-15  9:50                   ` Michael Shigorin
2008-10-15 12:31                     ` ahtoh
2008-10-15 12:35                       ` Led
2008-10-15 12:44                     ` Serge
2008-10-19 14:36                       ` Michael Shigorin

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=200810142021.28231.ahtoh@pulnet.ru \
    --to=ahtoh@pulnet.ru \
    --cc=sysadmins@lists.altlinux.org \
    /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

ALT Linux sysadmins discussion

This inbox may be cloned and mirrored by anyone:

	git clone --mirror http://lore.altlinux.org/sysadmins/0 sysadmins/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 sysadmins sysadmins/ http://lore.altlinux.org/sysadmins \
		sysadmins@lists.altlinux.org sysadmins@lists.altlinux.ru sysadmins@lists.altlinux.com
	public-inbox-index sysadmins

Example config snippet for mirrors.
Newsgroup available over NNTP:
	nntp://lore.altlinux.org/org.altlinux.lists.sysadmins


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git