I have an LSI 9300-16i 16-Port HBA card (Amazon: B0BYZBNXBS) sitting on an ASUS Pro WS W680M-ACE SE motherboard. Pre HBA firmware update, I was receiving a lot of error messages on boot. Post HBA firmware update, the errors went away, however, I recently started a ClamAV scan of my TrueNAS system with the --debug switch because I was receiving different errors in the ClamAV log. A day went by and I tried logging into my TrueNAS web UI and the web UI seemed to have a hard time connecting to TrueNAS. When the web UI finally allowed me to input my username/password, it took about a minute before a window popped up basically stating to make sure TrueNAS was connected to the network, of course which it is. I went to the text selection page and told TrueNAS to restart. TrueNAS restarts and then the HBA error messages start again. I powered down and from a cold boot, the error messages won’t go away and I can’t login to TrueNAS. So, apparently this HBA card is not working with my system as expected.
My motherboard BIOS is up to date with no issues and the firmware of the HBA card has been updated successfully with no issues. I did read the following from an HBA search - When adding an HBA card, it’s important to ensure that the card is set to IT (Initiator Target) mode rather than IR (Integrated RAID) mode, as IT mode provides raw drive access which is necessary for TrueNAS.
Could anyone recommend an HBA card that will actually work out of the box with a minimum of 8 drives, just plug and play?
@Johnny_Fartpants Well, my system is sitting powered off as I can’t get it to boot past the error messages, it’s literally hung on the error messages. I previously posted on this here:
Perhaps the card is overheating? It is known for that so people tend to strap cooling on it. Assuming it’s a real card not a knockoff, it’s a great card.
If it’s hung on error messages, post the screen. The previous post you had you were booting ok so something is different now.
@sfatula When I first started the machine after assembling it, the offending error message appeared on boot before the HBA card had time to get hot. As of this post, I opened a TrueNAS shell and ran a dmesg. The following repeating error message can be seen:
AER: Correctable error message received from 0000:01:00.0
pcieport 0000:01:00.0: PCIe Bus Error: severity=Correctable, type=Physical Layer, (Receiver ID)
pcieport 0000:01:00.0: device [10b5:8724] error status/mask=00000001/0000a000
pcieport 0000:01:00.0: [ 0] RxErr (First)
Nothing has changed to the system since updating the firmware on the installed HBA card. It was just sitting running.
I read the following post on the TrueNAS forum:
I don’t mind spending a little on a good HBA card, and from the above “HBA Selection” post, I just purchased a SAS9305-16i SATA SAS 16 Port HBA (Amazon: B0DY78R3HG). I need this to work as my current QNAP NAS is EoL and is exhibiting hardware issues. If this SAS9305-16i HBA doesn’t work, I’m at a loss.
So either the card has an issue or something else has an issue (sorry for stating the obvious). Sure getting a new card will help narrow things down but ideally I’d want to drop the card you already have into another system and see what happens. I’m sure I’ve heard about non legit cards but not sure exactly how you would tell them apart.
This guy is brilliant and I’d highly recommend him if you wanted to be sure you’re buying a legit card.
@etorix I looked at the 9300-8i, and yes, definitely for 8 drives. I chose the 9305 series because the “HBA Selection” post mentioned that the 9305-16i is a single controller, and uses less power. @Fastline mentioned in that post, “Yes, going for 9305 or 9400 as i have already tested it with my friend in TrueNAS Scale and it works fine.” As well, I could future proof my system in the event I purchase a full tower and deck it out with more than 8 drives.
@Johnny_Fartpants Unfortunately, I don’t have another machine to drop the card into. I’ll give the links a look see. Thanks for sharing. I’m not sure how I would check for a non legit card either. It’s just discouraging at the moment.
I’m supposed to get the SAS9305-16i HBA card on Sunday. Hopefully it will work.
@Johnny_Fartpants My boot drive is an 1 TB NVMe connected directly to the motherboard and is a Samsung 990 PRO (Amazon: B0BHJF2VRN). I have a second Samsung 990 PRO, a 2 TB, connected directly to the motherboard as well that I’m using for any TrueNAS app installations like ClamAV.
@etorix Looking at the picture of the SAS9305-16i that I just ordered, it doesn’t have a 6 pin power adapter like the HBA I presently have installed on the motherboard.
Let’s keep our fingers crossed and hope this new SAS9305-16i HBA works as expected. If it doesn’t, I don’t know…
@Johnny_Fartpants As I recall, somewhere in all of my searches, and I didn’t keep that search, showed that pcieport 0000:01:00.0 was the port the HBA is connected to. Somewhere I referenced it to Broadcom and found that the HBA was Broadcom. Something about PLX Technology PEX 8724 and the PEX 8724 is part of Broadcom’s range of PCIe switches and retimers.
From my “When Booting Several Error Messages Appear on Monitor” post, [10b5:8724] kept showing up in the errors as well as what I posted above. After I updated the firmware for that HBA card, a dmesg search showed the below:
sudo dmesg | grep -i 10b5
[ 0.285611] pci 0000:01:00.0: [10b5:8724] type 01 class 0x060400 PCIe Switch Upstream Port
[ 0.285897] pci 0000:02:00.0: [10b5:8724] type 01 class 0x060400 PCIe Switch Downstream Port
[ 0.286093] pci 0000:02:08.0: [10b5:8724] type 01 class 0x060400 PCIe Switch Downstream Port
[ 0.286285] pci 0000:02:09.0: [10b5:8724] type 01 class 0x060400 PCIe Switch Downstream Port
The HBA card that is in the computer has 4 connections to it. When I looked at the above dmesg grep search, there are 4 entries for that message and it didn’t show any errors at that time. I thought maybe the firmware update for the HBA card worked. Apparently not.
What would really be cool is when I replace the existing HBA with the new HBA, and connect all of the drives appropriately, there would be no errors on boot and all of my data would be accessible just like it should be. Now that would be something. I’m not holding my breath.
Indeed. A 9305 is fully powered from the PCI slot; a 9300-16i may require more power than the slot can deliver according to PCI specification, and has an additional power connector.
I restarted my system to see what might occur, and if it becomes corrupted, then it becomes corrupted. I figured if anything is going to happen, so be it. As a test, I enabled HTTPS using this method - System > General Settings > GUI > Settings button > checkmark box > Save
Which by the way doesn’t seem to work. My browsers won’t push to HTTPS for some reason; I even have my browsers set to only use HTTPS. After enabling HTTPS by checking the checkbox I restarted TrueNAS and interestingly enough I didn’t see the offending error messages as before. Odd. I then opened a shell and ran sudo dmesg
$ sudo dmesg | grep -i 10b5:8724
[ 0.278647] pci 0000:01:00.0: [10b5:8724] type 01 class 0x060400 PCIe Switch Upstream Port
[ 0.278957] pci 0000:02:00.0: [10b5:8724] type 01 class 0x060400 PCIe Switch Downstream Port
[ 0.279177] pci 0000:02:08.0: [10b5:8724] type 01 class 0x060400 PCIe Switch Downstream Port
[ 0.279388] pci 0000:02:09.0: [10b5:8724] type 01 class 0x060400 PCIe Switch Downstream Port
I just thought I would report what I’m seeing in the event anyone else is seeing the same. Maybe the TrueNAS developers might even see it. Anyway.
@Johnny_Fartpants The HBA that’s presently in the computer is x8 and will only fit in one PCIe slot on that motherboard. So yes, it’s in the correct lane, the x16. It won’t fit in any other slot on that board.
Amazon told me that the HBA is being delivered Monday instead of tomorrow. I’ll report back when I get it installed.
