Solved

Is RAID1 easier to recover from than RAID5?

Posted on 2011-03-24
8
638 Views
Last Modified: 2012-05-11
Hi, I am putting together a quote for a new Dell server. I am trying to decide whether to go with RAID1 or RAID5 -- I've done both setups in the past and also setups containing RAID1 for the operating system and RAID5 for data.

My basic question is--if a drive goes bad in RAID1, or furthermore if the disk controller goes bad but I have one drive that is OK, can I just boot to this single drive without needing to recreate the RAID1 LUN?

For this client they only need about 1-2 TB storage. I'm thinking of just using two 2TB drives in RAID1. This will host VMware ESXi (free version) and a couple vm's. I'm leaning towards using RAID1 because of its simplicity and also because it seems more reliable in a disaster scenario if the RAID controller goes out of wack. I just went through a horrible experience of a RAID5 LUN that crashed and I couldn't bring it back online. I don't think it was bad disks but a bad controller. I'm wondering if I would have been able to recover data if it was a RAID1 and I had one good disk--could I just pop it out of the broken server and then access on a different system?
0
Comment
Question by:goldylamont
[X]
Welcome to Experts Exchange

Add your voice to the tech community where 5M+ people just like you are talking about what matters.

  • Help others & share knowledge
  • Earn cash & points
  • Learn & ask questions
  • 2
  • 2
  • 2
  • +2
8 Comments
 
LVL 47

Assisted Solution

by:David
David earned 100 total points
ID: 35211618
The premium controllers will completely shield you from any problems, (assuming you pop for the battery backup for the cache on the controller).   The system won't even crash if a drive fails, or at least it would  have to be a multiple simultaneous failure scenario.

Go RAID1 if at all possible.

DO NOT get 2 x 2TB SATA disks, unless you look at the utilization.  Remember I/O is effectively RANDOM in a VMware environment.  Add all the overhead for filesystems and such and then compare how many RANDOM I/Os per second those disks run at, and you may very well have a system that "supports" only a few users.

0
 
LVL 33

Accepted Solution

by:
PowerEdgeTech earned 300 total points
ID: 35211620
"My basic question is--if a drive goes bad in RAID1, or furthermore if the disk controller goes bad but I have one drive that is OK, can I just boot to this single drive without needing to recreate the RAID1 LUN?" ... "Could I just pop it out of the broken server and then access on a different system?"

Yes you can, but:

You must put it in a system with a compatible controller (another PERC 6 or PERC H700).  When you do, the controller will see the configuration on the drive and mark it as "Foreign".  You will then Import that configuration into the controller.  As soon as you do, it will read and import the configuration on the disk and the data should be accessible (barring any data corruption).  Once you do this, you cannot put the disk back onto the PERC 6/i, as migration from the PERC 6 to H700 is supported, but backwards migration is not supported.

0
 
LVL 47

Expert Comment

by:David
ID: 35211678
You've most likely had all sorts of problems by buying consumer class SATA drives which are architecturally incompatible with most RAID controllers to begin with.  

Invest in quality storage designed for 24x7x365.  Either buy late-model fibre channel used, or SAS-2 disks.  I wouldn't limit to just RAID1, go with RAID6 or RAIDZ2.
0
Optimize your web performance

What's in the eBook?
- Full list of reasons for poor performance
- Ultimate measures to speed things up
- Primary web monitoring types
- KPIs you should be monitoring in order to increase your ROI

 

Author Comment

by:goldylamont
ID: 35211738
OK, so it sounds like both commenters are saying I'm correct in leaning towards RAID1. This great info so thanks.

dlethe I will look into RAID6 or RAIDZ2. I'm not sure where you are getting that I'm using or plan to use SATA drives though--I use SAS for vmware, and the drives that failed were SCSI (old server).

Thanks! If you have any other suggestions I'm all ears, otherwise I'll close out question.
0
 
LVL 33

Expert Comment

by:PowerEdgeTech
ID: 35211782
I didn't say it explicitely, but you inferred correctly ... I also prefer RAID 1.
0
 
LVL 27

Assisted Solution

by:Steve
Steve earned 50 total points
ID: 35212075
Raid 1: simple and quick data access. Easy to recover but wastes a lot of space.
Raid 5: more resilient and easier to expand. Slower data access.

Id recommend fast disks, 15k at least, and avoid SATA as they wont cope with what you want . go for SAS if you can afford it.

As mentioned above, the disk access in ESX is quite random and disks are usually the worst bottleneck on VM systems.

You'd get much better performance if you could afford 4 drives so you can create 2 x raid1 logical disks (disk 1 & disk 2 = logicaldrive1, disk 3 & disk 4 = logicaldrive2)
This would let you share the load between 2 sets of 2 disks and greatly reduce the bottleneck of all systems waiting for a single logical disk.
0
 
LVL 21

Assisted Solution

by:SelfGovern
SelfGovern earned 50 total points
ID: 35228039
OP, be careful of tonto's suggestions.

1) I don't know anyone else who'd say that RAID 5 is more resilient than RAID 1 -- in fact, in a 4, 6, or higher disk configuration, you can potentially lose two or more disks from a RAID 1+0 config without losing data, whereas losing two will always take down a RAID 5 setup.

If I had four disks, I'd put them in RAID 1+0, also called RAID 10, which is a combination of mirroring and striping.  Let the controller manage the physical disks... you worry about the data.   RAID 10 will potentially give you better performance, as I/O operations can get data from multiple disks at once.

If you want to partition your disk (so OS is on C, data on D, for instance) then by all means do that, but do it from a RAID 10, using the OS to split the disk, not the RAID controller.
0
 

Author Closing Comment

by:goldylamont
ID: 35369107
guys thanks. I'm going with RAID1 with two 600GB SAS drives.

dlethe, sorry for the misleading comment I made initially about going with two 2TB drives--I *thought* they were SAS drives on the Dell site but I was actually looking at "Near-line SAS" drives which are actually SATA drives with a SAS connection.

For other readers, my recommendation would be to use two SAS drives in RAID1 to host the VM operating systems. Then you could also add in two Near-line SAS/SATA drives for additional storage space (just for files/Office docs needing low disk I/O)
0

Featured Post

Building an interactive eFuture classroom

Watch and learn how ATEN provided a total control system solution including seamless switching matrix switch, HDBaseT extenders, PDU, lighting control to build an interactive eFuture classroom.

Question has a verified solution.

If you are experiencing a similar issue, please ask a related question

Is your phone running out of space to hold pictures?  This article will show you quick tips on how to solve this problem.
The business world is becoming increasingly integrated with tech. It’s not just for a select few anymore — but what about if you have a small business? It may be easier than you think to integrate technology into your small business, and it’s likely…
This tutorial will walk an individual through the process of installing of Data Protection Manager on a server running Windows Server 2012 R2, including the prerequisites. Microsoft .Net 3.5 is required. To install this feature, go to Server Manager…
Two types of users will appreciate AOMEI Backupper Pro: 1 - Those with PCIe drives (and haven't found cloning software that works on them). 2 - Those who want a fast clone of their boot drive (no re-boots needed) and it can clone your drive wh…

624 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question