Dell Avago LSI9305-16E not detecting arrays

[Greg0986]

Hi all,

I have acquired 2 LSI9305-16E's (Dell OEM) and I am having trouble getting storage arrays working with it. I have tried an older H200E with the array and it works fine, so we can rule that out. I have tried the 9305 in both Dell and HP servers (Dell R720 and R730, and HP Gen8 and Gen9) and it's detected by Windows and Ubuntu etc but the array doesn't appear. I have updated the firmware of the cards to the latest from Broadcom's website, and also did the bios while I was there. After updating the BIOS, I booted into the controller but that showed no devices attached.

The cables I am using to connect the array to the hba are SFF-8644 to SFF-8644, just one cable between the top controller in the array and the hba. I've tried different cables and made sure they are fully seated. Is this correct or do I need a diffeent cable or different configurations of cables?

Secure boot is disabled on the server I am working with. They are appearing in Ubuntu as LSI SAS3216 PCI-Express Fusion-MPT SAS-3.

The only thing that is suspect to me is the fact that the temp LED on the hba is flashing, however it's cool where the server is and the fans on the server are setup for increased cooling. I have also re-applied the thermal paste as it was old and crusty. Temps taken with a infrared thermometer on the heatsink shows 35c average.

Does anyone have any ideas why it isn't working? (apart from it saying its overheating)

 

[poto]

Not sure if you resolved your issue, but I just returned same model hba as defective, identical issues.

Dell OEM 9305-16e would flash temp fault led red upon power-up. Megaraid Storage Manager reported idle temp 60C, well within operating range. IR thermometer on heatsink read 54C. Additional fan lowered temp further, but temp fault led remained lit.

Allowed for possibility of anomaly between Dell hba and SuperMicro motherboard, updated firmware & bios to latest. Connected to disk shelf, but no devices detected. Also tried different port configurations in bios setup, no difference.

Possible previous over-temp damaged IOC & fault led is "stuck" on.

 

[Greg0986]

I did what no one should do and resolved the issue without updating this thread!

I had to wipe the flash, using "sas3flash.efi -o -e 7" in DOS mode
Then boot into efi mode, and run "sas3flash.efi -f SAS9305_16e_IT.bin -b mptsas3.rom -b mpt3x64.rom"

Without wiping the flash memory, the new firmware wouldn't work.

 

[poto]

Did the erase & reflash resolve all the issues? My hba was showing red temp fault before I updated any firmware or bios.

 

[Greg0986]

Yes it did, before flashing it was coming up with the red led temp fault, I assume it was something to do with the system it was connected to before.

 

[poto]

After using the erase function on region "7", did you have to re-write SAS address and/or device ID, or just reboot to efi mode & apply command you listed? I've not had to reflash in some time, & want to avoid bricking this!

"sas3flash.efi -f SAS9305_16e_IT.bin -b mptsas3.rom -b mpt3x64.rom"

 

[Greg0986]

I didn't re-write the SAS address, it hasn't caused any issues so far. I just wiped the flash, and then rebooted into efi mode and ran that command. Obviously make sure the files are named that and are in the same folder where the sas3flash.efi file is.

I roughly followed this guide: https://www.servethehome.com/flash-lsi-sas-3008-hba-e-g-ibm-m1215-mode/

 

[poto]

Thanks, I'll do a bit of a refresh before proceeding.

 

[poto]

The mode 7 erase & re-flash corrected the temp fault led, and arrays are now visible. I'm 2hrs into stress-testing, and no discrepancies. Documentation on this model is sparse, but it seems the temp fault led is multi-purpose. Thanks for following up on your solution - it saved an unneccessary return.

 

[zackiv31]

Sorry to revive an old thread, I think I got one of these cards (Dell OEM) and the red/green light are on immediately when I turn on the server. Doesn't show in `lspci` in ubuntu. `sas3flash.efi -list` under UEFI shows No Avago SAS adapters found (Version 13.00.00.00). Been reading the forums and am kind of stuck. Anything simple I may be missing?

 

[movax]

Blast from the past but this thread helped me out yesterday -- had a 9305-16e (Dell Branded) that would not link up with drives. The red fault LED was blinking + dmesg was spamming "over-temp" messages (at 40 C... which is pretty reasonable!).

I had to make a DOS bootable stick to use sasflash to erase (didn't work from Linux), erased Zone 7 (all flash) after 'backing up' the image + writing down the SAS address / board-specific stuff -- problem solved, fault went away!

Sorry to revive an old thread, I think I got one of these cards (Dell OEM) and the red/green light are on immediately when I turn on the server. Doesn't show in `lspci` in ubuntu. `sas3flash.efi -list` under UEFI shows No Avago SAS adapters found (Version 13.00.00.00). Been reading the forums and am kind of stuck. Anything simple I may be missing?

View attachment 34053

If it doesn't show up in lspci, I think the card is likely toasted. When I did the full erase above, the card had no firmware and would not show up in sas3flash either, without selecting it by PCI address. The silicon should still enumerate on PCIe + report VID/PID, it just has no Fusion-MPT firmware on-board.

 

[bobn4burton]

If it doesn't show up in lspci, I think the card is likely toasted. When I did the full erase above, the card had no firmware and would not show up in sas3flash either, without selecting it by PCI address. The silicon should still enumerate on PCIe + report VID/PID, it just has no Fusion-MPT firmware on-board.

So I'm considering trying this. I just got a used Dell 9305-16e off ebay and I think I'm seeing similar symptoms. I can actually see the card show up in Windows 11 Device Manager. But I don't see any SATA HDDs after connecting them.
I noticed that there are two LEDs on the HBA. Red LED1 which is used for thermal status (I believe). Green LED2 which is used for a heartbeat. I noticed that immediately after power on, the green LED2 heartbeat starts blinking about once every second. But the red LED1 thermal LED also comes on solid. Then after a few seconds (seems to correspond to when POST finishes and windows starts booting), the green LED2 just shuts off. And the red LED1 thermal LED starts blinking. Maybe once a second or so...and just keeps blinking.

Does this sound similar? And if so...what other info do I need. I was thinking it would be as simple as using the sasflash utility (I tried using the .exe version from within Windows 11 command prompt) to list the card and get it's info. But your comment about it not showing up after deleting the flash has me a bit concerned. If it doesn't show up anymore after erasing the flash, how do I re-flash it with updated FW? You mentioned something about using the PCI address??? Any tips on that part?

 

[bobn4burton]

OK...I fixed my issue. For anyone else that has a similar issue, I'll post my findings here (and also in the another thread that I started initially to ask for some help for finding this one).

Ultimately it was the blinking thermal LED as my best clue combined with this thread that helped me fix the issue. So if anyone is troubleshooting issues, I'd definitely recommend checking the thermal and/or heartbeat LED as part of the first steps in troubleshooting.

I tried updating from the current v14 FW that came on my 9305-16e LSI card to the latest version on Broadcom website which is version 16.00.12.00. The FW update went just fine...but didn't solve the issue. Interestingly, just this FW update alone DID actually change the behavior slightly on my LEDs.
Originally my LED behavior consisted of:
For the first 10-15 seconds after power on, my thermal LED stayed solid red and my heartbeat LED would flash green.
Then after the first 10-15 seconds, my thermal LED would start flashing red and the heartbeat LED would just turn off completely.

After the FW update to version 16, I had high hopes for the first few seconds because the first power on after update and my heartbeat LED was blinking green but no thermal red LED at all. But after about the same 10-15 seconds the green heartbeat LED would just shut off completely and the thermal LED would start blinking red.

So the FW update alone didn't fix the issue.

I then did a flash erase. I tried the flash erase with option 6 first (not 7 as suggested in this thread). Please note that for the 9305-16e, there is actually an errata noted in the knowledge base that an erase with option 6 actually wipes out 3 of the card specific numbers as well. The tracer number, assembly number and SAS address number. But if you just make note of those first (they should also be printed on stickers on your card), you can reprogram them pretty simply afterward. Here is the article with details: Flashing SAS9305-16e with ‘–o –e 6’ parameters erases some controller values

After doing that erase (note I only did an erase level 6, not 7), reprogramming the lost 3 numbers and then redownloading the latest v16 FW for the board...everything was fixed. No more thermal red led at all. And my green heartbeat LED was blinking as expected...both immediately after power on and also after the 10-15sec time period where it would normally change behavior.

And after that, all of my drives showed up perfectly and everything is going smooth so far!!

HUGE thanks to Greg0986 for finding the solution and taking the time to post the solution!! Saved me a bunch of hassle on a card return...