Dynamic Smart Array B120 RAID construction

[Frank Sauerbier]

Hi all,

does anyone know how the RAID 1+0 is constructed with a Dynamic Smart Array Controller B120i?

Is it 1122 or 1212 regarding Bays 1234?

Thanks!

Frank

[Paul_J_K]

Hello @Frank Sauerbier,

I hope the following answers your query along with the diagramatic illustration. 

When the array contains only two physical drives, the fault-tolerance method is known as RAID 1.When the array has more than two physical drives, drives are mirrored in pairs, and the fault-tolerance method is known as RAID 1+0 or RAID 10.

In each mirrored pair, the physical drive that is not busy answering other requests answers any read requests that are sent to the array. This behavior is called load balancing. If a physical drive fails, the remaining drive in the mirrored pair can still provide all the necessary data. Several drives in the array can fail without incurring data loss, as long as no two failed drives belong to the same mirrored pair.This fault-tolerance method is useful when high performance and data protection are more important than the cost of physical drives.

Advantages:
•This method has the second highest read performance of any fault-tolerant configuration.
•No data is lost when a drive fails, as long as no failed drive is mirrored to another failed drive.
•Up to half of the physical drives in the array can fail.

Disadvantages:
•This method is expensive, because many drives are needed for fault tolerance.
•Only half of the total drive capacity is usable for data storage.

Please login to the SSA and select the B120i controller.
If you are yet to configure, all 4 drives should be in unassigned state.
Go into unassigned drives and select all 4.
Use the "Create Array" option and choose RAID 1+0 for the RAID level along with other parameters. "Finish".

You can check the drive assignment map to see which drives belongs to the mirror. Ideally in your words,

11 - in RAID 0 with - 22; where 11 and actually bay 1 and 2 and 22 is actually bay 3 and 4. This may vary due to other configuration variables. So best thing is to check the big drive assignment map in the ADU report.

[Frank Sauerbier]

Hi Paul,

thank you so much. I learned the magic is the "big drive assignment map" and the "paired drive" table.
In my situation I got drives from a crashed RAID 1+0 in a ML350 Gen8 with B120i (MS 2012 Essentials).
I was said:

at first the HDD in Bay 1 failed and was exchanged.
Reconstruction started automatically. But it was not possible to monitor any progress in SSA (may be firmware to old (from 2012)). After one day they rebooted the server and updated BIOS, ilo and B120i.
During this (?) the disk in Bay 3 failed and was exchanged (ready for rebuild).
Then "mirror split" (????) and start rebuild manually.
Two logical drives were rebuilded ....
After all no boot possible, not to RAID or DVD (..illegal opcode ...).

So I got the 4 old and the 2 exchange-drives.
I was able to place the 4 old drives in a Microserver Gen8 and got 2 logical drives.
ADU here "Paired Drive" LD1 says:
0 Physical Drive (1 TB SATA HDD) 1I:1 Physical Drive (1 TB SATA HDD) 3I:3 Informational
1 Physical Drive (1 TB SATA HDD) 2I:2 Physical Drive (1 TB SATA HDD) 4I:4 Informational
2 Physical Drive (1 TB SATA HDD) 3I:3 Physical Drive (1 TB SATA HDD) 1I:1 Informational
3 Physical Drive (1 TB SATA HDD) 4I:4 Physical Drive (1 TB SATA HDD) 2I:2 Informational

The strange thing is: Disk in Bay 2 seems to be complete empty. Disks in Bays 1,3,4 have data.
The two new exchanged drives (are HPE, all others HP) seems to be empty also.

When I access these two LD (with testdisk) partition table error is reported and many files are missing .....

I think my only chance is to find a working RAID0-Pair in correct disk order. In all recovery software they ask for block/stripe size and offset. Do you know what the corresponding numbers in ADU report are?

Thanks a lot.

[Paul_J_K]

Hi @Frank Sauerbier ,

You might consider checking areas like the following in the ADU report.

Logical Drive 1 : Logical Drive Identification Block Size 512 Bytes Per Block (0x0200)

In your configuration, 1 & 3 are mirrored and so are 2 & 4. 

Also note that being an entry level controller, the B120i uses driver based RAID algorithm at the software level. So rebuilding and many other activities require the OS driver in action. 

[Frank Sauerbier]

Hi Paul,

thanks. Does this mean that 1&3 is the first "disk" in RAID 0?

So I can check pairs 1+2, 1+4, 3+2 and 3+4?

Is RAID information written to every disk?

Can I check if this information is consistent and equal on every disk?

And I need to understand what the logical volume means. My LD2 has a data offset 0x09fff600.

160MB? 160GB could be the size of LD1.

Other Volume Signature 0x00000000
OS 0xfd00
Controller Order 0x00
Additional Info 0x00
Data Offset 0x09fff600
Backed Out Write Drives 0 (0x00)
Parity Distribution Mode 0x00
Parity Shift Factor 0x04

[Paul_J_K]

Hi @Frank Sauerbier ,

Yes, 1&3 appears to be a mirrored pair which means the data on 1 and 3 are same. 

Yes, RAID information (RIS) resides on the member disks of the RAID. 

Checking further in terms of the data consistency on each disk is not something which I know about since this is managed by the RAID algorithm and not available as readily readable data.

Being a software dominated RAID algorithm, it is not suggested to swap around disks other than the original configuration and also the OS (RAID driver) running has a big part to play in this type of (Dynamic Smart Array) RAID.

[Frank Sauerbier]

Hi Paul,

thanks again.

I am going on to understand things ....

At the moment I am confused about Block Size / Stripe-Size.

When I build an array I have the following data:

Strip-Size/Full Strip-Size: 256KiB/512KiB

In the ADUReport this data is not referenced .... block size is given as 512 Bytes (512 Bytes Per Block (0x0200))

How may I understand this?

The data recovery software wants to have the block size and they said this is normally identical to stripe size.