[CentOS] Advice on CentOS 7, software raid 1, lvm, 3 TB

Wed Sep 17 17:26:29 UTC 2014
Jeff Boyce <jboyce at meridianenv.com>

Sorry for breaking the threading, as I only get the daily digest.  My 
comments (interspersed) begin with the **.

Message: 41
Date: Tue, 16 Sep 2014 18:38:02 -0400
From: SilverTip257 <silvertip257 at gmail.com>
To: CentOS mailing list <centos at centos.org>
Subject: Re: [CentOS] Advice on CentOS 7, software raid 1, lvm, 3 TB
<CA+W-zoqD9e826==2H6i9g4CkM_SrqW=yEtFcWntAy=qeaZGdLA at mail.gmail.com>
Content-Type: text/plain; charset=UTF-8

On Tue, Sep 16, 2014 at 6:07 PM, Jeff Boyce <jboyce at meridianenv.com> wrote:

> Greetings -
> I am preparing to order a new desktop system for work. In general the new
> system will be a Dell Precision T3610 with two 3 TB drives. I plan on
> installing CentOS 7 as a KVM host, with virtual machines for Win 7 Pro and
> Linux Mint. I am looking for some advice or a good how-to on configuring
> software raid on the two drives, then using LVM for the host and virtual
> machines. I have configured our company server (Dell T610) with hardware
> raid and LVM for a CentOS 6 KVM host and several virtual machines so I am
> not a complete novice, but have never setup a Linux software raid system,
> and have not played with a CentOS 7 install yet. I have been searching the
> web and forums for information and am not finding much for good guidance.
> Lots of gotcha's are popping up identifying issues related to CentOS 7,
> software raid 1, grub install, > 2 TB disks (or any combination of these
> factors). The CentOS Wiki has a good description of installing

I'm not sure which wiki article you might have read. That URL might be
worthwhile to share.

** The CentOS Wiki article I was referring to was the same one you provided 
in the first link of the group of references posted at the bottom of your 
note.  I was a little put off by the article having fairly significant 
warnings in the first two paragraphs of the article, so I only skimmed 
through it.

> CentOS 5 with raid 1, but there is a big warning about being an
> unsupported (risky) approach. Can anyone point me to a good how-to, or
> provide some general guidance. Thanks.
Hopefully what I have typed up below helps you.
I don't know about soft-raid1 being an unsupported/risky approach ... that
said I'd pick hardware raid over software raid (considering I had spare
hardware) so I don't have to fuss with raid at the OS level. I have worked
on a mix of software-raid and hardware-raid systems (and still do) ... each
has its own pros/cons. I've had success re-adding a new drive in degraded
soft-raid1 arrays in a production environment ... so I say go for it.

[ ]
Somebody else asked about C7 and soft-raid in the past week or week and a
You can find that thread here:
Though I don't think much was accomplished in that thread.

** Yea, I saw that thread recently also, and was hoping for some good 
information from it, but the original post wasn't specific enough to 
generate specific guidance.  That is why I thought to try a post that was 
more specific to my objective.

My suggestion to you (as well as that last person) is to spin up a VM (or
spare bare-metal hardware) and use mdadm commands to assemble, stop,
hot-fail, hot-remove, and rebuild (add a new disk to replace a "failed
one") your soft-raid array.

** I am planning on having some time set aside to play/experiment with the 
new box while setting it up.  So I am fortunate to not be in a situation 
where I have to get it into production now before figuring it all out.

As is the case with many things Linux, the manpage is your friend.
Sometimes sysadmins and hobbyists decide to publish what they've done
(good or bad) which can be found with the search engine of your choice.

** I am a book worm so I don't mind reading man pages.  But as an ecologist, 
it is often good to understand the big picture before diving into the 
details.  That way I understand which detail I need to look at, and what 
order all the details go together in.  It looks like the ArchLinux links you 
gave me below will give me an understanding of that as I read them in 

In this case, even generic (non-RH or non-CentOS specific) command
documentation is likely what you want. More than likely you'll get the
results you want by booting to a rescue CD (or switching to a shell on your
install CD), setting up your soft-raid, then booting to your install CD,
which will probe for your disk/soft-raid/lvm layout.

steps for graphical approach (C5, so dated) -
partitionable soft-raid -
TLDP create soft-raid - http://www.tldp.org/HOWTO/Software-RAID-HOWTO.html

** I've seen these first three, and will go back and look at the first one 
in more detail now as mentioned above.

Arch Linux create soft-raid -

** I had not found this one.  My first impression scanning through it is 
that it looks to have a lot of the other subjects I am interested in (mbr 
vs. gpt, XFS filesystem, EUFI vs BIOS mode).

repairing - http://www.thomas-krenn.com/en/wiki/Mdadm_recovery_and_resync
repairing - http://ethertubes.com/repair-md-mdadm-raid/
repairing -

** Yes, thanks for these also.  I am planning on outlining and testing my 
disaster recovery process when I set this up so I don't have to figure it 
out in a crisis when/if it happens.

**  Thanks for your time.

// SilverTip257 //

Jeff Boyce