Dynamic smart array loses logical drives - G8 / B120i

[slowtraveler]

Dear collegues,

I have problem with my home Microserver G8  (Bios J6) with Dynamic smart array controller B120i (.v.3.54.0)

System is 'stock' Celleron with following drives attached:

Bay 1&2 - 2TB HDD in RAID 1 

Bay 5 - 120GB SSD (bootable - OS)

OS - FEDORA 30

I create Logical drives  in two arrays

- A - SSD (booting)

- B - RAID 1 for both HDD

Everything seems working OK when setting it up. Restarting, rebooting, powering off soon after setting it up - no problem.

Both logical drives are OK during POST(power on self test), and system boots flawlessly...

But when I try to reboot after longer period - a day or two (or month), POST during SMART ARRAY B120i controller initiallisation reports 0 logical drives (due to 'oflline drive' etc...). And logically - server cannot boot.

When I get to HP Storage administrator - arrays are not present, but physical drives are there.

I create Logical drives/arrays back, and everything is starting as usual.

Until next occasion.. This problem makes UPS powering off (upon UPS battery depletion) very dangerous, because in fact server will be powered back when electric power is back, but most probably it will not boot, and will need console operation to recover it. .

Can you give me some hint where can be the problem?

[slowtraveler]

Am i the only with such problem? Or i explaned the problem not clearly?

Any ideas are welcome..

Is B120i have some memory battery, which can be depleted and controller loses it's memory?

Or OS write there?

[Jazz_ISS]

Hello,

Are you using HPE Genuine drives or not?
Are the failing drive always in the same bay/slot?Or have you tested with drives in different slots as well?
Ensure to check the power and data cables  for any loose connectivity.

Do rememebr the following points for HP Dynamic Smart Array B120i Controller:-
HDD bays 1 and 2 support 6.0Gb/s SATA.
HDD bays 3 and 4 support 3.0Gb/s SATA.
NOTE: Supports RAID 0/1/10 only. No FBWC support.

If issue remain unresolved, I would recommend log a support ticket with HPE so that we can check whether there are Hard Read/write errors on the installed drives which in turn could be the reason for logical drives intermittent failure.

[slowtraveler]

Thanks for reply...

To questions:

1. No.. SSD is 128GB (ADATA SP550), Raid 1 array is 2x2TB WD Blue serie equial drives (WDC WD20EZRZ-00Z)

2. I cannot see failing physical drive, but just logical arrays are missing... Because when i start HP Administrator during boot (selecting F5 during B120i initialisation), i can see all physical drives and to re-create logical areas without any problem. Controller loses both logical arrays at that time - a single bootable SSD (connected to internal SATA plug), and 2x2TB HDD RAID 1 array on bay 1&2.. So i must re-create arrays and to select bootable array too, and after that server boots as usual...

3. Checked, uninstalled, installed drives in bay - no change...

P.S. thinking if subject HDD disk needs more spool-up time? But controller loses SDD array too, but not only RAID1 array (reports 0 (zero) logical arrays found. And everythime HP Administrator founds the drives and i can re-create arrays flawlessly.. 

P.P.S. how i can log support ticket to HPE - it need contract number i suppose? Is there support to such old server?

[Jazz_ISS]

Hello,

Yes to log a support case with HPE you need to valid Support Contract ID . But do remember HPE will not participate in Log Analysis if in case Non-HPE parts are in use.

[Jimmy Vance]

---

@slowtraveler wrote:

Thanks for reply...

To questions:

1. No.. SSD is 128GB (ADATA SP550), Raid 1 array is 2x2TB WD Blue serie equial drives (WDC WD20EZRZ-00Z)

2. I cannot see failing physical drive, but just logical arrays are missing... Because when i start HP Administrator during boot (selecting F5 during B120i initialisation), i can see all physical drives and to re-create logical areas without any problem. Controller loses both logical arrays at that time - a single bootable SSD (connected to internal SATA plug), and 2x2TB HDD RAID 1 array on bay 1&2.. So i must re-create arrays and to select bootable array too, and after that server boots as usual...

3. Checked, uninstalled, installed drives in bay - no change...

P.S. thinking if subject HDD disk needs more spool-up time? But controller loses SDD array too, but not only RAID1 array (reports 0 (zero) logical arrays found. And everythime HP Administrator founds the drives and i can re-create arrays flawlessly.. 

P.P.S. how i can log support ticket to HPE - it need contract number i suppose? Is there support to such old server?

---

Being this controller is the embedded chipset plain SATA controller the ODD port is tied to the controller.  The RAID function was an add on to support the special RAID driver for a few specific operating systems. The ODD port was never tested or in the design spec of the system to support anything other than a CD/DVD. While many have succesfully made what you are attempting to do work, it is not a supported configuration.

[slowtraveler · Accepted solution]

Yes, i know it's 'not supported' configuration, but it's strange that configuration works well when restarted soon after setup, and loses 'low level' configuration when not restarted soon. So i'm curious is it hardware fail, or just effect of unsupported devices.. 

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

P.S. Finally after few unsuccessful experiments i switched off built-in controller B120i, and moved to standard Legacy mode.

Server want's to boot from Bay 1  HDD (maybe - lowest accessible bay), but my SDD with bootable OS is installed on 'bay5'. So in BIOS i select in ' boot order ' - 'second SATA controller', as bootable  (evidently 'first SATA controller' controls bay 1-4, and in that case it tries to boot from bay-installed HDD, but not SSD).

So, the case is closed...