Hi Folks,

 

 

***Update**

 

A solution/workaround found see below, this applies to latest Unraid 6 as well, THE Disabling IRQ issue affects my system where it comes down to a snails pace once "Disabling IRQ # occurs

 

As mentioned in the subject I'm experiencing issues with virtualized unraid and IRQ #18 message which brings everything to a screeching halt of snails pace.

 

 

The build is very similar to JohnM's Atlas build:

http://lime-technology.com/forum/index.php?topic=14695.0

 

My Build:

• Intel Xeon E3-1230
  SUPERMICRO MBD-X9SCM-F-O LGA 1155 Intel C204
  Kingston 16GB 2x (2 x 4GB) 240-Pin DDR3 SDRAM ECC Unbuffered
  3 X SUPERMICRO AOC-SASLP-MV8
  
• 
  

 

In terms of the confirguration I followed Johnm's guide to the T and everything seems to work fine but however I get Disabling IRQ error, it appears random does not seem to be related to running 3x SUPERMICRO AOC-SASLP-MV8, I Ran 2 and had experienced the same issue.

 

Below is extract from syslog.txt:

Mar  4 15:38:53 MediaTower kernel: irq 18: nobody cared (try booting with the "irqpoll" option)

Mar  4 15:38:53 MediaTower kernel: Pid: 0, comm: swapper Tainted: G        W  2.6.32.9-unRAID #8

Mar  4 15:38:53 MediaTower kernel: Call Trace:

Mar  4 15:38:53 MediaTower kernel:  [<c10451cf>] __report_bad_irq+0x2e/0x6f

Mar  4 15:38:53 MediaTower kernel:  [<c1045305>] note_interrupt+0xf5/0x13c

Mar  4 15:38:53 MediaTower kernel:  [<c1045a14>] handle_fasteoi_irq+0x5f/0x9d

Mar  4 15:38:53 MediaTower kernel:  [<c1004a82>] handle_irq+0x1a/0x24

Mar  4 15:38:53 MediaTower kernel:  [<c1004285>] do_IRQ+0x40/0x96

Mar  4 15:38:53 MediaTower kernel:  [<c1002f29>] common_interrupt+0x29/0x30

Mar  4 15:38:53 MediaTower kernel:  [<c1008160>] ? default_idle+0x2d/0x42

Mar  4 15:38:53 MediaTower kernel:  [<c1001a14>] cpu_idle+0x3a/0x4e

Mar  4 15:38:53 MediaTower kernel:  [<c128a8bf>] rest_init+0x53/0x55

Mar  4 15:38:53 MediaTower kernel:  [<c13f580c>] start_kernel+0x27b/0x280

Mar  4 15:38:53 MediaTower kernel:  [<c13f5091>] i386_start_kernel+0x91/0x96

Mar  4 15:38:53 MediaTower kernel: handlers:

Mar  4 15:38:53 MediaTower kernel: [<c11ead61>] (usb_hcd_irq+0x0/0x5b)

Mar  4 15:38:53 MediaTower kernel: [<f84bf24e>] (mvs_interrupt+0x0/0x66 [mvsas])

Mar  4 15:38:53 MediaTower kernel: Disabling IRQ #18

 

Full Syslog:

http://www.sobon.ca/Unraid/syslog.txt

 

 

My cat /proc/interrupts:

 

My Syslinux.cfg with irqfixup and irqpoll added in attempt to remedy the issue with no luck:

 

default menu.c32

menu title Lime Technology LLC

prompt 0

timeout 50

label unRAID OS

  menu default

  kernel bzimage

  append initrd=bzroot irqfixup rootdelay=10 irqpoll

label Memtest86+

  kernel memtest

 

 

Any help would be appreciated, keep in mind this is a virtualized unraid not sure what bios changes are possible.

 

I managed to find a workaround/Solution, keep in mind this is not a root cause fix, but a better known fix aside from rebooting the box:

 

Adding "noirqdebug" into the syslinux.cfg file:

default /syslinux/menu.c32

menu title Lime Technology

prompt 0

timeout 50

noirqdebug

label unRAID OS

  menu default

  kernel /bzimage

  append initrd=/bzroot

label unRAID OS Safe Mode (no plugins)

  kernel /bzimage

  append initrd=/bzroot unraidsafemode

label Memtest86+

  kernel /memtest

Try with either irqfixup OR irqpoll but not both at the same time.  You can see if there is a BIOS update available also.  Disable anything not used.

 

I had a similar issue on my ESXi build but irqpoll seems to have fixed mine.

 

You can try running a beta version instead of 4.7 and that might help/fix the issue.  I am running 5.0b6a (soon will be upgraded most likely) and irqpoll works.

Try with either irqfixup OR irqpoll but not both at the same time.  You can see if there is a BIOS update available also.  Disable anything not used.

 

I had a similar issue on my ESXi build but irqpoll seems to have fixed mine.

 

You can try running a beta version instead of 4.7 and that might help/fix the issue.  I am running 5.0b6a (soon will be upgraded most likely) and irqpoll works.

 

Thanks, I'm going to try the Irqpoll standalone see if that helps, in terms of bios I have the latest installed.

Update, Irqpoll did not fix the issue sil getting disabling IRQ #. Any suggestions?

Im having a similar problem on my new openfiler box. its killing my fiber card.

I have your exact setup minus one MV8.  I will check tonight and see if I see the same error but I do not believe it is happening.  It certainly isn't affecting speeds any if I do have the same errors in my log.  15 2TB drives currently.

It almost looks like a glitch or bug on syslinux, I was able to find this issue outside of the unraid forums as well.

 

Does anyone know what exacly triggers this? I can't seem to pinpoint it and it's getting annoying to reboot on a daily basis to use.

I upgraded to the latest beta hoping that would fix the issue however it has not, unraid worked fine for about 30 hours and then I got his with Disabling IRQ# again. My patience is running out, there has to be a solution for this.

I'll look in my past logs. as far as i know I have not seen these.

It looks like I was able to partially solved the IRQ issue, I used the "noirqdebug" option which doesn't let the debug disable IRQ's however this does not really point to why it's happening if anyone has any other ideas let me know.

 

I actually got the errors you described in OP.  For me I just restarted the VM and the really slow access went away and I didn't have any of the IRQ#18 errors in the new log.  Hoping it will always be that easy to clear.

I seem to be getting a similar error, documented here: http://lime-technology.com/forum/index.php?topic=23417.msg206539#msg206539

 

Using a SASLP-MV8 and I only noticed this sporadically on shutdowns. When I do see "Disabling IRQ #16" message, the subsequent reboot I always lose the same disk on the MV8. Extremely weird.

 

X8SIL-F, X3430

M1015 on pci-e slot #2

MV8 on pci-e slot #3

Also work on Unraid 6.x, note some updates overwrite the syslinux.cfg file which will need to be reapplied.