Z800 - Embedded LSI SAS RAID Controller - FAIL Boot 90% of Time on RAID 0 With New BX500 SSD Drives

[lifefork]

Hello,

From what I've seen from google searches, even though they are old, this is the best place to look for answers for this old workstation. I've had it a very long time, and plan to keep it for at least a few more  years, so I upgraded the RAID array I had (RAID 5 with SAS drives) to RAID 0 with 4 "new" crucial BX500 drives. 

Odd things happened during installation, for example the optical drive wouldn't load the Windows 10 installation media, or even wouldn't boot at all. Works fine otherwise. I ended up making a USB drive and installing through USB 2 port.

Anyway, I went through installation, multiple reboots, and everything went ok until it was done. Reboots after this point *most of the time* end up with the controller telling me the volume has failed. Specifically:

Array Boot FAIL

So, I go into the LSI config, and everything looks just fine...every single time. Everything is OK.

I haven't found any rhyme or reason to what makes it boot. I can go into the LSI config, maybe it will boot then, maybe into BIOS settings then it boots after that, this time I just fully powered down and unplugged and it booted on startup.

Are there any Z800 pros left on the forum? Has anyone run into this. I'm at a loss right now.

I *believe* this is a revision 3 motherboard. Funny, the last number of 00x is warped/half-rubbed off. Looks like a 3 maybe.

2 x X5670

96GB Memory

ASUS RX570

Startech USB3 card

Bluetooth dongle

4 x Crucial BX500 drives in RAID 0 array on the SAS controller.

850W

Any help is appreciated!

Kind Regards

[DGroves]

several things you need to be aware of

1. the onboard LSI SAS/SATA controller like the z800's onboard SATA controller has a limit on bootable partition size it must be under 2TB in size if using the latest bios, non bootable drives can be 3TB drives

2. using 4 sata ssd's in raid zero is pointless speed wise as the sata hardware is limited to approx. 580 megabytes per second which a single sata ssd can achieve

4. while it is possible to upgrade the onboard LSI firmware it's not recommended doing as any mistake will render the LSI part non working (and possibly the entire system)

a better solution is to install a adaptec ASR-6805T raid card, which is a  PCI-e express 2.0 card (same as z800 has)  this card will upgrade the z800's onboard sata 2 into  sata 3 drive speeds and the card will support larger non bootable drive sizes up to 4TB (or any size if using a later OS with GPT partitioning support like win 10/11)

https://www.ebay.com/itm/195577450315?epid=28031091139&itmmeta=01J43MGC6BBCGEGXNA82V05308&hash=item2...

and this card unlike most LSI cards supports Raid and JBOD modes and SSD's, just make sure to get the card with the onboard battery backup/cache card, these cards are quite cheap on ebay nowadays (or the newer ASR-71605 "PCI-e 3.0" if you plan on using it in a newer system someday)

https://storage.microsemi.com/en-us/support/raid/sas_raid/sas-6805/

next you will need at least one "Forward" Mini SAS SFF-8087 cable for the 6805 card

next you will need 4 SATA male to male adapters to allow the 6805 cards drive connectors to mate with the z800's 4 internal drive bay cables

https://www.ebay.com/itm/313702967275?itmmeta=01J43MQNGF2337C9ZZC1G1WBCJ&hash=item490a2767eb:g:oIQAA...

[lifefork]

Thanks for your response, I really appreciate it.

The reason I'm still using this server, and also the drives, is that  for the past 15 years I have just salvaged the parts I want for my own person systems, not really going for the best things as I can sell those easier. So, I have these matching ssd drives, and can get them combined into a larger drive. 

I did decide to go and follow your instructions to  prepare for possible future upgrades, and got an Adaptec ASR-71605, following your suggestions for the cabling, etc.

Unfortunately, the Adaptec card is not detected in any PCIE slots. It lights up, the batter works, but no go. I reset BIOS defaults on the machine. I will have to dig out another machine when I get the time, and see if the card itself is the culprit.

Else, is there something on the machine side I should be doing? 

Kind Regards

[DGroves]

disable the z800's onboard LSI bios from within the  z800's bios ( will disable booting from onboard controller)

then disable the LSI device itself to prevent windows  from seeing the onboard controller

power off, if using dual cpu's ( which you have) select any  free 8x/16x pci-e slot, if using single cpu white pci-e slots are  disabled

card will sign on during boot and prompt for CTRL-A to enter adaptec bios

[lifefork]

I have disabled the 1068E controller inside of the LSI utility, and also disabled "sas" controller under security tab in the z800 bios.

I have even just continually hit ctrl-a through the startup process just in case the display was hidden for some reason or another. I can't see the adaptec device.

[lifefork]

Today, I pulled the 71605 and put it into another machine to see if it pops up. Not detected, windows does not detect.

Placed the 71605 back into the z800 and hooked the bugged raid 0 array back up onto the list reenabling the lsi boot and controller. Windows did not see the 71605 either when I finally was able to get the array to work.

I believe the card is bad, unless anyone has anything else to day. @DGroves thank you for your support.

*** On another note, since I have reset the bios settings, the fans run at full speed constantly again. Somehow I turned that down years ago.

[DGroves]

the raid card may be defective, when installed and powered on do the red lights on the card sweep back and forth? if not bad card

the 71605 is a pci-e 3.0 card and is backward compatible with the z800's pci-e 2.0 bus

[lifefork]

Yes, correct. Solid red. In process of exchanging! Will have to wait for two transit times.

[DGroves]

default case fan speed is set from the bios, note that if this setting has no effect the ambient temp sensor located in the power cable has most likely failed and you can either replace the cable or locate the sensor and just replace it

the z800/z820/z840 use a transistor that is embedded in the power on cable for measuring the ambient air temperature when this fails (or is missing) the cpu fans will ramp up to full speed

the solution is to either replace the cable, or replace the bad transistor in your existing cable

ebay cable assembly:

https://www.ebay.com/itm/332629885597

if  replacing the sensor in the cable, it's near the power button board end close to where the speaker wires exit the black heat shrink that covers the cable wires , and it's a common npn transistor you can use a 2n3904 or a 2n222a plastic body transistor as a replacement

note that the existing transistor will have two leads connected together and a single lead connected to colored wires,  match the existing leads  to the replacement and also note which lead connects to which color wire

https://components101.com/transistors/2n2222a-pinout-equivalent-datasheet

[lifefork]

Thanks for your response on this. I had a couple days ago ran across a post of yours talking about this very thing.

Ambient temp 0 in BIOS.

Ordered an inexpensive pack of transistors and will follow your advice. Extremely helpful.

Kind Regards