Questions tagged [irq]
42 questions
2
votes
1 answer
Can't update smp_affinity
As part of the experiments that we are running to optimize the performance of the applications on our Linux box we are trying to assign all the network interrupts to a single cpu core.
Unfortunately, it doesn't seem to work the way it should. E.g.…
![](../../users/profiles/192534.webp)
sad2k
- 21
- 2
2
votes
0 answers
Disabling IRQ#16 when removing DVI cable
I have a computer running with a Blackmagic Decklink card and a usbscreen. The problem i'm having is somethimes a irq conflict occurs and the OS disables irq16 witch both the decklink card and the usb screen is connected to. If i remove the dvi…
![](../../users/profiles/144110.webp)
Ryezs
- 21
- 1
2
votes
0 answers
Adaptec 5805 on Centos 6 IRQ 16 error
I'm having a problem with Adaptec 5805 raid card
http://www.adaptec.com/en-us/support/raid/sas_raid/sas-5805/
(with two SAS discs in raid) and Gigabyte motherboard…
![](../../users/profiles/104356.webp)
Radek
- 153
- 2
- 7
2
votes
3 answers
VMWare ESXi and IRQ balancing
VMWare ESXi 4.1
Guest: Linux kernel 2.6.32 64bit (tried older as well)
In multi-CPU virtual machine all interrupts (eth0,ata_piix) are stuck to the single CPU. When running on the pure hardware, everything is balanced ok.
Tried E1000 network driver…
![](../../users/profiles/6080.webp)
disserman
- 1,850
- 2
- 17
- 35
2
votes
1 answer
Spikes on softirqs RCU monitored with NetData on proxmox container
I have a server with the following characteristics: https://www.soyoustart.com/it/offerte/1801sysgame05.xml
Processor
Intel i7-4790K
RAM
32GB DDR3 1333MHz
Traffic
Unlimited
Anti-DDoS
Included
Disks
1x240GB SSD
Bandwidth
250 Mbps
I've installed…
![](../../users/profiles/190006.webp)
Joseph
- 121
- 3
1
vote
0 answers
too many nic intterupts caused performance issue on haproxy server
i have a ubuntu 18.04.1 server running with 8-core processor and 8 Gigabyte's of memory. it's a cloud server which is based on KVM virtualization. i am using haproxy 1.8.8 for load balancing on my servers. problem is when i run load-test on my…
![](../../users/profiles/546157.webp)
aliopalopina
- 11
- 1
1
vote
1 answer
Setting /proc/irq/100/smp_affinity is not keeping
I am setting the smp_affinity mask on my two NIC IRQs (100 and 101) to a specific cpu core - echo 0002 > smp_affinity - but the new value only stays for a couple seconds then starts randomly cycling through all the cpu cores again. Is there another…
![](../../users/profiles/182920.webp)
Michael Martinez
- 2,543
- 3
- 20
- 31
1
vote
1 answer
Setting affinity for network interrupts in Windows Server 2012 R2
We have an application that sends and receives a lot of network traffic - but interrupts are causing us to get lower than max performance. At 20 Mbps input and 20 Mbps output (on a 1 Gbps line), the network interrupts use 100% of core 0 and the…
![](../../users/profiles/300997.webp)
akonskarm
- 13
- 1
- 4
1
vote
1 answer
Setting default_smp_affinity on Linux to the correct value
I have 12 core boxes:
egrep -c processor /proc/cpuinfo
12
cat /proc/irq/default_smp_affinity
fff
I would like to use all of the cores so my understanding is that the default_smp_affinity should be set to ffffffffffff.
echo ffffffffffff >…
![](../../users/profiles/5258.webp)
Istvan
- 2,562
- 3
- 20
- 28
1
vote
0 answers
IRQ not handled by all CPUs
I have a problem about balancing irq. I took irq numbers from /etc/interrupts and assigned each irq to each of 24 CPUs as this tutorial says
for i in {143..166};do cat /proc/irq/${i}/smp_affinity; done
000001 …
![](../../users/profiles/92682.webp)
ibrahim
- 431
- 7
- 20
1
vote
0 answers
XCP Project Kronos syslog error: "irq ... : nobody cared" on Dom0 host
One of our production clusters driven by XCP suddenly went uresponsive. After restart and some investigation we found such logs in dom0 machine syslog:
Oct 26 20:32:03 hetzner-2-mrx kernel: [1797931.659040] irq 339: nobody cared (try booting with…
![](../../users/profiles/142929.webp)
Vlad Fedin
- 111
- 1
1
vote
0 answers
Solaris 10 / NIC Marvell Yukon 88E8001 hungs with latest driver
I've installed latest yukon driver from marvell for Solaris 10 8/11. Got the driver, installed, updated /etc/driver_aliases and worked fine.
After the system is running for a couple of days, it hungs with the following error:
Mar 24 12:13:33…
![](../../users/profiles/115417.webp)
gustavo
- 11
- 1
1
vote
1 answer
"No irq handler for vector (irq -1)" after hibernation
Resuming from hibernation messages like these are printed to all terminals:
Message from syslogd@Ildjarn at Apr 8 08:11:14 ...
kernel:[24089.310964] do_IRQ: 0.82 No irq handler for vector (irq -1)
Message from syslogd@Ildjarn at Apr 8 08:16:14…
![](../../users/profiles/76420.webp)
Kasreyn
- 71
- 1
- 7
1
vote
1 answer
Apparent IRQ conflict driving me nuts under CentOS
The Environment
I have a small, dual-core Intel Atom-based server running CentOS 5.5 x64 with a slightly customized Xen kernel. It also has one on-board 10/100 NIC with an additional 3-port 10/100 NIC. Within this server, I also run a single Xen…
![](../../users/profiles/30996.webp)
jamieb
- 3,387
- 4
- 24
- 36
1
vote
1 answer
What does the irqstat graph in munin represent?
What does it mean for this munin graph of weekly irqstat to suddenly go crazy like it does -- is it bad? What are local timer interrupts?
![](../../users/profiles/3842.webp)
Silas Snider
- 148
- 5