CentOS

Hey guys,

I just setup a server using CentOS 7.5.1804, and I've selected a Marvell SATA RAID controller (Marvell 88SE9230), running two SATA drives at RAID 1.

The storage system appears to be functioning as expected, but I don't know how to interrogate the RAID status, trigger rebuilds, or monitor the underlying drive status. Some Googling suggests that smartctl is the proper interface, but I'm having poor results so far. Hoping you can guide me.

Code: Select all

# smartctl -a /dev/sdb -T permissive
smartctl 6.5 2016-05-07 r4318 [x86_64-linux-3.10.0-862.9.1.el7.x86_64] (local build)
Copyright (C) 2002-16, Bruce Allen, Christian Franke, www.smartmontools.org

=== START OF INFORMATION SECTION ===
Device Model:     MARVELL Raid VD
Serial Number:    676bccacc2c10010
Firmware Version: MV.R00-0
User Capacity:    4,000,703,119,360 bytes [4.00 TB]
Sector Size:      512 bytes logical/physical
Device is:        Not in smartctl database [for details use: -P showall]
ATA Version is:   ATA/ATAPI-7 T13/1532D revision 4a
Local Time is:    Tue Jul 24 10:08:59 2018 PDT
SMART support is: Unavailable - device lacks SMART capability.

Any recommendations on how to manage this RAID device?

Thanks!

Using FakeRAID is almost always a recipe for disaster at some point. The Redhat knowledge base article about such controllers recommends using them in AHCI mode and using linux software RAID instead. Reading man smartctl shows that there is a -d marvell switch but I have no idea if that will work with your particular controller.

TrevorH wrote: ↑
2018/07/25 12:08:14
Using FakeRAID is almost always a recipe for disaster at some point.

Thanks for the response!

So, until this comment, I wan't aware of the term "FakeRAID" or the differentiation from actual RAID controllers. Thanks for the introduction.

Any recommendations on low-cost SATA RAID controllers that are well-supported by CentOS?

RAID-1 does not require significant compute power; software on CPU is fine for it and pure software RAID is less dependent than fakeRAID (which is essentially software RAID).

The drives are probably directly visible and mdadm can manage the fakeRAID arrays. (Not sure if true for Marvell.)

Code: Select all

lsblk
cat /proc/mdstat

If you go for RAID-6, where discrete processor, cache RAM, and battery backup do make a difference, then you won't be in the realm of "low number of drives" any more.

CentOS

Marvell SATA RAID on 7.5.1804

Marvell SATA RAID on 7.5.1804

Re: Marvell SATA RAID on 7.5.1804

Re: Marvell SATA RAID on 7.5.1804

Re: Marvell SATA RAID on 7.5.1804