AOC-SAS2LP-MV8


Recommended Posts

Excellent news guys.

 

thanks for the updates!

 

Where do you guys get them from so others can get their grubby little fingers on them.

Cheapest appears to be Provantage.  I've ordered 3 IPMI cards for my X7SBE boards from WiredZone they have them too for a little more and in Stock.  My search turned up these others.  When you include shipping WiredZone was actually the cheapest for me.
Link to comment
  • Replies 53
  • Created
  • Last Reply

Top Posters In This Topic

Wow, the pro vantage price is the same for the sas2lp-mv8, as most other places price for the saslp-mv8.

 

Seriously considering buying a couple of these to work in my Norco 4224 with all Hitachi 3TB drives. Question is now, what are the backplanes rated at, SATA II or SATA III?

 

Sas2. So sata3 works

Link to comment

I was considering purchasing the AOC-SAS2LP-MV8 for my unRAID 4.7 server, but didn't know if it was supported that far back or if I'd need to upgrade to the beta 5, which I'm not really prepared to do yet.

 

It looks like it is only has drivers in 5beta12(a) at this point. It is still not really "supported" at this point.

I would guess that in the future it would be, but time will tell.

Link to comment
  • 2 months later...

I have just built a system with three AOC-SAS2LP-MV8s and have serious issues with it on 5.13 and 5.14. However, I suspect a card problem as my WHS system with one card also has issues. On WHS writes fail within few seconds and the drive disappears from Windows. This only happens in 64-bit windows, not 32-bit.

 

On unRAID multiple cards do not work well together. The system (see specs below) stalls when the card BIOS boots are enabled. The Marvell BIOS only recognises two cards and the boot halts at the second Marvell BIOS screen. This happens when:

  • Two cards are installed in the two 8x slots and a hard drive is connected to ports 2-7 on the second card
  • Three cards are installed and any hard disk is connected to Adapter 1 (as reported by the BIOS)

 

So I disabled the adapters BIOS boots in the board BIOS. unRAID was able to boot and the three cards were recognized. However,

 

Issue 1: When drives are connected to any adapter the following errors are logged:

---------------------------------------------------------------------------------------------

Nov 23 19:16:42 SS kernel: ata7: sas eh calling libata port error handler (Errors)

Nov 23 19:16:42 SS kernel: ata8: sas eh calling libata port error handler (Errors)

 

Issue 2: IRQ 16 disabled

----------------------------

Most of the time I am also getting an IRQ 16 disabled error. If this happens the section below appear three times (one for each adapter?).

The error occurs most of the time especially when drives are connected to adapter 0, ports 5, 1 ports 3-6 or 2, ports 2-7.

 

The PCI report shows IRQ 16 is used by two cards (one uses IRQ 17). IRQ 16 also seems to create the slow write speeds to the drives (see next issue).

 

Nov 23 18:28:56 SS kernel:  sdc: sdc1

Nov 23 18:28:56 SS kernel: sd 8:0:0:0: [sdc] Attached SCSI disk

Nov 23 18:28:56 SS kernel: drivers/scsi/mvsas/mv_sas.c 1205:phy 0 attach dev info is 0

Nov 23 18:28:56 SS kernel: drivers/scsi/mvsas/mv_sas.c 1207:phy 0 attach sas addr is 0

Nov 23 18:28:56 SS kernel: drivers/scsi/mvsas/mv_sas.c 1205:phy 1 attach dev info is 0

Nov 23 18:28:56 SS kernel: drivers/scsi/mvsas/mv_sas.c 1207:phy 1 attach sas addr is 0

Nov 23 18:28:56 SS kernel: drivers/scsi/mvsas/mv_sas.c 1205:phy 2 attach dev info is 0

Nov 23 18:28:56 SS kernel: drivers/scsi/mvsas/mv_sas.c 1207:phy 2 attach sas addr is 0

Nov 23 18:28:56 SS kernel: drivers/scsi/mvsas/mv_sas.c 1205:phy 3 attach dev info is 0

Nov 23 18:28:56 SS kernel: drivers/scsi/mvsas/mv_sas.c 1207:phy 3 attach sas addr is 0

Nov 23 18:28:56 SS kernel: drivers/scsi/mvsas/mv_sas.c 1205:phy 4 attach dev info is 0

Nov 23 18:28:56 SS kernel: drivers/scsi/mvsas/mv_sas.c 1207:phy 4 attach sas addr is 0

Nov 23 18:28:56 SS kernel: drivers/scsi/mvsas/mv_sas.c 1205:phy 5 attach dev info is 0

Nov 23 18:28:56 SS kernel: drivers/scsi/mvsas/mv_sas.c 1207:phy 5 attach sas addr is 0

Nov 23 18:28:56 SS kernel: drivers/scsi/mvsas/mv_sas.c 1205:phy 6 attach dev info is 0

Nov 23 18:28:56 SS kernel: drivers/scsi/mvsas/mv_sas.c 1207:phy 6 attach sas addr is 0

Nov 23 18:28:56 SS kernel: drivers/scsi/mvsas/mv_sas.c 1205:phy 7 attach dev info is 0

Nov 23 18:28:56 SS kernel: drivers/scsi/mvsas/mv_sas.c 1207:phy 7 attach sas addr is 0

Nov 23 18:28:56 SS kernel: scsi9 : mvsas

Nov 23 18:28:56 SS kernel: irq 16: nobody cared (try booting with the "irqpoll" option)

Nov 23 18:28:56 SS kernel: Pid: 0, comm: swapper Not tainted 3.1.0-unRAID #2

Nov 23 18:28:56 SS kernel: Call Trace:

Nov 23 18:28:56 SS kernel:  [<c104fa80>] __report_bad_irq+0x1f/0x95

Nov 23 18:28:56 SS kernel:  [<c104fc2d>] note_interrupt+0x137/0x1a8

Nov 23 18:28:56 SS kernel:  [<c104e76a>] handle_irq_event_percpu+0xef/0x100

Nov 23 18:28:56 SS kernel:  [<c1050146>] ? handle_edge_irq+0xcb/0xcb

Nov 23 18:28:56 SS kernel:  [<c104e79f>] handle_irq_event+0x24/0x3b

Nov 23 18:28:56 SS kernel:  [<c1050146>] ? handle_edge_irq+0xcb/0xcb

Nov 23 18:28:56 SS kernel:  [<c10501af>] handle_fasteoi_irq+0x69/0x82

Nov 23 18:28:56 SS kernel:  <IRQ>  [<c1003566>] ? do_IRQ+0x37/0x90

Nov 23 18:28:56 SS kernel:  [<c130bea9>] ? common_interrupt+0x29/0x30

Nov 23 18:28:56 SS kernel:  [<c11dd951>] ? acpi_idle_enter_bm+0x22a/0x25e

Nov 23 18:28:56 SS kernel:  [<c1272de0>] ? cpuidle_idle_call+0x75/0xbd

Nov 23 18:28:56 SS kernel:  [<c1001a5f>] ? cpu_idle+0x39/0x5a

Nov 23 18:28:56 SS kernel:  [<c12fb578>] ? rest_init+0x58/0x5a

Nov 23 18:28:56 SS kernel:  [<c145172d>] ? start_kernel+0x28c/0x291

Nov 23 18:28:56 SS kernel:  [<c14510b0>] ? i386_start_kernel+0xb0/0xb7

Nov 23 18:28:56 SS kernel: handlers:

Nov 23 18:28:56 SS kernel: [<c12435ee>] usb_hcd_irq

Nov 23 18:28:56 SS kernel: [<f84730c9>] mvs_interrupt

Nov 23 18:28:56 SS kernel: [<f84730c9>] mvs_interrupt

Nov 23 18:28:56 SS kernel: Disabling IRQ #16

 

 

Issues 3: Generally low write speeds and a few writes fail

-----------------------------------------------------------------

  • Drives connected to Adapters 0 (the 4x slot) and 2 ports 2-7 are 1-3 Mb/s
  • Drives connected to adapter 1 and 2 port 0+1 are about 50 Mb/s
  • I am also getting 50 Mb/s with drives connected to the motherboard
  • In a few cases the write did fail altogether. I always wrote directly to the drive.
  • I thought the slow write speeds were related to the IRQ16 but in case of the adapater 1, 3+4 the IRQ16 error was present but write speeds were ok
  • I haved tried changing various BIOS setting including PCI latency and PCI Gen1 force with out any difference.

 

Supermicro has acknowledged the WHS issue and and is currently working with Marvell on a solution. The unRAID issues is reported but not yet acknowledged by Supermicro.

 

Welcome any recommendations from the community on what else I could try to circumvent or better describe/isolate the problem.

 

tazman

 

 

Link to comment

run memtest if you have not, just to rule that out.

 

add the irqpoll option to you syslinux.cfg file on the flash drive.

 

I added irqfixup to mine for a similar reason:

root@Andromeda:/etc/ntp# cat /boot/syslinux.cfg                                                                              
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                                                                                 
label FreeDOS                                                                                                                
  kernel memdisk                                                                                                             
  append iso initrd=fdbasecd.iso                                                                                             
label Memtest86+                                                                                                             
  kernel memtest

Link to comment

I was considering purchasing the AOC-SAS2LP-MV8 for my unRAID 4.7 server, but didn't know if it was supported that far back or if I'd need to upgrade to the beta 5, which I'm not really prepared to do yet.

 

It works fine in 4.7

 

he is questioning the the AOC-SAS2LP-MV8.

 

To my knowledge, they are reported as not working with any version prior to 5beta12 or 11?

 

The AOC-SASLP-MV8 works in all supported versions of unraid except 5Beta12-12a

Link to comment

I just completed an extensive memory test which was ok. The irqpoll and irqfix options did not fiex the "IRQ 16 disabled" message.

 

In the meantime Supermicro has acknowledged the problem and is working with Marvell on a resolution. The boot stall with multiple cards is a know issue and the resolution is the disabling of the PCI adapter BIOS boots for all cards except the one with the boot drive on.

Link to comment

Can someone explain to me why this model is not shown on Supermicro's website? Moreover I can't find it in my country (even though Supermicro is not well distributed, it's strange). What are the differences with the previous model (the previous one is available here in Switzerland)?

Link to comment

The difference is SAS1 vs SAS2 standards.. the biggest is 3GB/s (On PCIe 4x) Vs 6GB/s (on PCIe 8x)

 

 

The SAS1 is fine for unraid.

HOWEVER, the SAS1 card is over-saturated when all drives are transferring data after about the 6th drive. Therefor slowing down things like parity checks and rebuilds.

The extra time however is  not a huge amount and still a solid card for unraid. As you might have noticed the SAS1 works in all supported version of unRAID with the exception of beta12/12a. while the SAS2 is only working only in the latest betas.

Link to comment

HOWEVER, the SAS1 card is over-saturated when all drives are transferring data after about the 6th drive. Therefor slowing down things like parity checks and rebuilds.

The extra time however is  not a huge amount and still a solid card for unraid.

 

I question if this is a real world condition, or just an issue of numbers and theoretical limits.

 

How many drives (i.e. unRAID class candidates)  go above 120MB/s ?

 

I'm running 5 drives on 2 AOC-SATA PCI-X controllers Both on the same PCI-X 133Mhz bus.

I'm getting 70,000 to 80,000 Kb/s parity sync at the start.

It slows down a bit as it drags on, but probably because I'm running 10 of the older EACS 5400 1TB WD Green drives.

Link to comment

The difference is SAS1 vs SAS2 standards.. the biggest is 3GB/s (On PCIe 4x) Vs 6GB/s (on PCIe 8x)

 

 

The SAS1 is fine for unraid.

HOWEVER, the SAS1 card is over-saturated when all drives are transferring data after about the 6th drive. Therefor slowing down things like parity checks and rebuilds.

The extra time however is  not a huge amount and still a solid card for unraid. As you might have noticed the SAS1 works in all supported version of unRAID with the exception of beta12/12a. while the SAS2 is only working only in the latest betas.

 

That's weird, I've had great success running 3 x SAS1 cards (2 fully populated, one with 6 drives) 12 and 12a...

Link to comment
  • 3 months later...
  • 1 month later...

Anybody using one of these with a Supermicro C2SEE motherboard?

 

Assuming it works with the above, given the option to purchase one of these vs another card, is there any reason not to choose this card over an LSI based card like the below?

 

AOC-USAS2-L8i

IBM M1015

 

 

Link to comment
  • 1 year later...

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.