Im cant find any info on what might be causing this error. Should i try booting irqpoll option? if so how do i do that. From what i can tell is it something with the motherboard or maybe a BIOS setting. Before i upgraded from 6.3.2 i had the issue a few times but it never caused any issues i dont think.

 

 

 

[Here is the log snapshot](http://imgur.com/a/kT21H)

 

[Motherboard](http://www.dfi.com.tw/products/product.html?productId=461)

 

[SAS Card](https://www.amazon.com/gp/product/B005B0Z2I4/ref=oh_aui_detailpage_o04_s00?ie=UTF8&psc=1)

 

 

Apr  5 21:24:29 Atlantis kernel: mdcmd (41): spindown 0

Apr  5 21:24:30 Atlantis kernel: mdcmd (42): spindown 29
Apr  5 22:28:20 Atlantis kernel: mdcmd (43): spindown 1
Apr  5 22:28:21 Atlantis kernel: mdcmd (44): spindown 2
Apr  5 22:33:17 Atlantis emhttp: cmd: /usr/local/emhttp/plugins/dynamix/scripts/tail_log syslog
Apr  5 22:36:34 Atlantis kernel: irq 16: nobody cared (try booting with the "irqpoll" option)
Apr  5 22:36:34 Atlantis kernel: CPU: 4 PID: 0 Comm: swapper/4 Not tainted 4.9.19-unRAID #1
Apr  5 22:36:34 Atlantis kernel: Hardware name: To be filled by O.E.M. To be filled by O.E.M./SB630-CRM, BIOS 4.6.5 04/30/2013
Apr  5 22:36:34 Atlantis kernel: ffff88061e303ec8 ffffffff813a3bf2 ffff8806004fd200 ffff8806004fd200
Apr  5 22:36:34 Atlantis kernel: ffff88061e303ef0 ffffffff810863e1 ffff8806004fd200 0000000000000000
Apr  5 22:36:34 Atlantis kernel: 0000000000000010 ffff88061e303f28 ffffffff810866aa ffff8806004fd200
Apr  5 22:36:34 Atlantis kernel: Call Trace:
Apr  5 22:36:34 Atlantis kernel: <IRQ> 
Apr  5 22:36:34 Atlantis kernel: [<ffffffff813a3bf2>] dump_stack+0x61/0x7e
Apr  5 22:36:34 Atlantis kernel: [<ffffffff810863e1>] __report_bad_irq+0x2b/0xb4
Apr  5 22:36:34 Atlantis kernel: [<ffffffff810866aa>] note_interrupt+0x1a0/0x22e
Apr  5 22:36:34 Atlantis kernel: [<ffffffff81084294>] handle_irq_event_percpu+0x3d/0x46
Apr  5 22:36:34 Atlantis kernel: [<ffffffff810842d3>] handle_irq_event+0x36/0x54
Apr  5 22:36:34 Atlantis kernel: [<ffffffff810871f6>] handle_fasteoi_irq+0x90/0xf8
Apr  5 22:36:34 Atlantis kernel: [<ffffffff81020515>] handle_irq+0x17/0x1b
Apr  5 22:36:34 Atlantis kernel: [<ffffffff8101ff6a>] do_IRQ+0x46/0xc2
Apr  5 22:36:34 Atlantis kernel: [<ffffffff8167e942>] common_interrupt+0x82/0x82
Apr  5 22:36:34 Atlantis kernel: <EOI> 
Apr  5 22:36:34 Atlantis kernel: [<ffffffff81551f5a>] ? cpuidle_enter_state+0xfe/0x156
Apr  5 22:36:34 Atlantis kernel: [<ffffffff81551fd4>] cpuidle_enter+0x12/0x14
Apr  5 22:36:34 Atlantis kernel: [<ffffffff8107c453>] call_cpuidle+0x33/0x35
Apr  5 22:36:34 Atlantis kernel: [<ffffffff8107c635>] cpu_startup_entry+0x13a/0x1b2
Apr  5 22:36:34 Atlantis kernel: [<ffffffff81035432>] start_secondary+0xf5/0xf8
Apr  5 22:36:34 Atlantis kernel: handlers:
Apr  5 22:36:34 Atlantis kernel: [<ffffffff814f95ff>] usb_hcd_irq
Apr  5 22:36:34 Atlantis kernel: [<ffffffffa03bf60a>] mvs_interrupt [mvsas]
Apr  5 22:36:34 Atlantis kernel: Disabling IRQ #16

 

 

 

Thanks for your time and help. 
 

 

atlantis-diagnostics-20170406-0128.zip

Irq used by the SAS2LP is getting disabled crippling its performance, check for board bios update and/or use a different pcie slot if available.

Irq used by the SAS2LP is getting disabled crippling its performance, check for board bios update and/or use a different pcie slot if available.

 

So what you are saying is its my motherboard that is disabling my sas card but any idea why? I've only got one pci e on it and the bios is up to date.

 

Linux is disabling the IRQ used by the controller, not the controller itself.

What exactly does that mean? This is all new to me especially the Linux side of things.

 

 

I can't tell you why it's happening, only that it is happening, try the workarounds I listed above, if none of them work replace the controller or wait for a new release, sometimes a newer kernel is enough to make it stop.

So you say sometimes a new kernel can cause the issue. I believe the issue only appeared after 6.3.3. is there a was I can roll back to 6.3.2 if so how?

I do have a different motherboard​ I could swap in and try that would be last resort tho.

Copy the 3 bz* files from v6.3.2 to the flash share and reboot.

Awesome, so I am back running 6.3.2 I suppose​ now only time will tell. Ive seen the error before but I don't thinks it's ever caused my speeds to drop.

If I keep getting the error would a board swap be the next best attempt?

5 hours ago, johnnie.black said:

check for board bios update and/or use a different pcie slot if available.

 

If none of those work, swap board or probably better, the controller, get one LSI based.

Well that's a bummer because I got the SAS card about a month ago and I haven't seen this issue. Is there an LSI JBOD card you recommend. I don't need raid functionality or anything like that.

 

Now you said it could be a kernel issue if that's the case will switching back to 6.3.2 give me a better idea?

 

Are supermicro cards known to have issues or something?

Yes Supermicro cards, specifically the SASLP and SAS2LP have been known to be problematic with unRAID. I had one in my system and it was ok for awhile then I started having all kinds of issues with it. I eventually replaced it with a Dell Perc H310 flashed to IT mode, resolved my issues.

Edited April 6, 20179 yr by ashman70

24 minutes ago, sourCream said:

Are supermicro cards known to have issues or something?

 

Not Supermicro controllers in general, but the SASLP and SAS2LP, although most users don't have issues with them there's more than a few with issues and unRAID v6, so controllers with LSI chipset are recommended (Supermicro also has some of those, though Dell H200/310 and IBM M1015 tend to be the cheapest available).

Are there any that come pre flashed in IT mode? Would you possibly have a link to one that would have that and support HDD spin down?

There is a guide on here on how to flash it, its really simple. I imagine you might be able to find one pre flashed, not really sure you would have to look.

The ones I mentioned above when crossflashed to IT mode become an LSI 9210/9211-8i, you can also by those LSI models, although they are generally considerably more expensive, all support spin down.

So something like this should work just fine then?
http://www.ebay.com/itm/112327269957

Yep

Awesome well I'll get it and give it a shot. I may as well use this card over the Sas2lp if it works with unRAID better any way. I guess then if I still have the issue it's got to be the board no?

I've only seen this behavior (Linux disabling the controller IRQ) with the SASLP/SAS2LP, both use the same driver, never saw it with an LSI controller.

Well I suppose that is good news in a way. Johnnie i really appreciate the help with all of this. I'll be sure to get that card and let you know what the outcome is. I overpaid for that SAS2LP but live and learn I'll be able to make most back if this fixes my issue. Also with IT mode I take it that is basically JBOD ? As in it reads drives just as they are no raid? And it should just be a matter of swapping the cards?

 

 

3 minutes ago, sourCream said:

Also with IT mode I take it that is basically JBOD ? As in it reads drives just as they are no raid? And it should just be a matter of swapping the cards?

 

Correct.

So I have a quick question I found this one that has IT mode flashed but in the description it says it's flashed with P19 and you can ask for P20 do I need to worry about that at all? This one has a year warranty that's why I ask about this one vs the other one

Both work fine (btw, there was a problem with initial P20 release, 20.00.07.00 works fine)

This is the one I found sorry I forgot to link it http://www.ebay.com/itm/291641245650