DL380p Gen8 P420i 1783-Slot 0 Drive Array Controller Failure

[AdamRA4]

I have an HP DL380p Gen8 server with all current firmware installed on everything (BIOS, RAID, iLO, etc).  This server has sat in a Datacenter from day 1 but has had an issue when we first bought it brand new a few years ago.  The issue being it would start up normally, boot into WIndows 2012 R2 fine but if we had to reboot, it would stick at the BIOS loading screen "Power and Thermal Calibration Complete".  Rarely it would get past that and start booting in to Windows but would freeze at the WIndows Loading screen.  Fans would then crank up to 50% and just sit like that until we iLO'd in and would do a hard shut off and turn it back on.  It was always in production so we never had the change to fully troubleshoot it or contact HP about warranty claims (dumb I know).  We have since pulled this from the datacenter and wanted to have it in house as a test server for various things so we ordered some extra hard drives that were the same as the ones already in it (653960-001 2.5" 300GB 15k SAS).  When we installed these into slots 5-8 we got the error Slot 0 HP Smart Array P420i Controller nitalizing... 1783-Slot 0 Drive Array Controller Failure! [Self-test failure (cmd=0h, err=00h, lockup=014:0h)].  So we pulled all the drives out, unplugged the server, waited 5 minutes, plugged all back in, same error.  We could not launch the HP Smart Array manager with this error.  We took out the original drive and put the new ones in slots 1-4 and it booted fine, we could access Smary Array Manager, it saw all 4 drives, we could create an array, installed Windows, and it booted.  We then took the original drives and plugged them into slots 5-8 along side the new drives in slots 1-4, we got the failure error again.  We unplugged everything again, removed the FBWC card, unplugged the battery from it, let it sit for a while, plugged it all back up, same error with all 8 drives plugged in.  I messed around with different drive orders, different amounts of drives, etc and I would always get this same error.  I did eventually get it to see 5 drives if I used slots 1,2,3,4,6 only.  I could create an array with those 5 drives, install Windows, etc.  Also it didnt matter which physical drives they were (new or old and mixed).  Max I could get was 5 using those specific slots.  I tried some consumer SSD's laying around in case it was a power issue but they were never detected in any slot.  There is no issue with the left bay (slots 1-4) no matter what drives I use, I only have a problem with the right bay (slots 5-8).  Specs of the server Dual E5-2620, 64GB memory, dual 460W PSU, DVD-RW, no add in cards, stock 4 port GB LAN adapter, and the hard drives I have listed. All genuine HP parts, no 3rd party parts.  I have attached pictures with all 8 drives giving me the error and a picture with no FBWC.  I dont want to use an add-in RAID card as the fans spin up and are loud in our little office.  With no add-in cards, they run around 6%-7% and is almost silent.  Since its the integrated RAID card, I cant swap it unless I board swap.  Any ideas?  Do I need to upgrade power supplies to dual 750W?

[Jimmy Vance]

As for the hanging during POST, try resetting NVRAM using SW6.

The other issue could be the drive backplane or the cables between the controller and the backplane, re-seat the cables   If you have a P420/P822/P830 card you could at least test to see if it is an issue with the embedded controller.

[AdamRA4]

Would resetting the BIOS back to defaults be the exact same as SW6 switch or are they different? I just reset the BIOS, rebooted a few times and everytime it gets stuck at the same part no matter how long I let it sit (I have let it sit for over 2 days once!). I also tried the SW6 switch reset (unplug, switch, plug back in turn on, wait for msg, power off, unplug, reset switch, power back on) and I get the same results.

I have swapped the cables (plug port 1 in 2 and vice versa) as well as reseated them. Even with the cables swapped, I only had a problem on the right storage bays (slots 5-8).

I read an older post about a Gen7 doing something similar and they replaced the FBWC card and it fixed it for that one specific person. Everyone else said updating the BIOS fixed it for them, but I already have the latest BIOS.

I dont mind ordering a part to replace a faulty one, I just dont want to go down the rabbit hole replacing multiple parts with no resolution.  I do not have any exyernal RAID cards to try it out with unfortunately.

[Jimmy Vance]

SW6 does a deeper reset than the RBSU option.  Since swapping the cables doesn't change anything, sounds like an issue with the backplane.  If you don't have any cables connected to the controller, does POST still hang?

[AdamRA4]

It only seems to hang when I reboot from Windows.  If I reboot before Windows (iLO config, Array config, BIOS, etc) it restarts just fine.  I will disconnect the cables and install Windows on an SD Card.