Tell me more ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.
up vote 2 down vote favorite

I'm a relatively inexperienced server admin who primarily works with Windows and a little bit of Linux. I'm the only admin of a small organization.

I'm looking for a way to have a high-availability storage cluster that will allow me to easily add more storage (new servers) to the cluster in the future. These servers would also need to replicate data throughout the cluster automatically and appear as one entity to the rest of the network. I'm looking for file-level storage, not necessarily block-level. And this is over a Gigabit Ethernet network.

I've heard of FreeNAS, though from what I can tell, it isn't designed for a HA cluster situation. I've also looked at Windows Storage Server with Failover, but the whitepaper on that says each server must have access to a shared storage array, whereas I'm looking for each server in the cluster to have its own array.

What solutions do you know of that provide basic high-availability storage clustering, and can you point me to any resources that can help me get started? Can you point me to any resources that explain the process of adding a new server to the cluster?

EDIT:

My budget is relatively low - not including hardware, maybe $1000. Ideally I'd build my own server with the specs I need and then install software on it to facilitate clustering. Overall I hope to spend $12,000 or less per 50 TB.

share|improve this question
What's your budget? – SpacemanSpiff May 21 '12 at 21:15
@SpacemanSpiff see edit. – Phil May 21 '12 at 21:21

3 Answers

up vote 2 down vote

Lot's of options here, but subjective to budgetary and design requirements. Off the top of my head you could look at:

1) Microsoft Distributed File system, required an AD infrastructure most likely.

2) Clustered NAS, look at a dual head NetApp system, very versatile, hard on the pocketbook.

3) Roll your own... use some open source tools like you mentioned and setup replicated storage.

share|improve this answer
up vote 1 down vote

OpenFiler can do that, it's free, but it's pretty complicated to setup (HA cluster, not the system itself). You can buy support from them though.

Other solution is clustered filesystems, for example Oracle Cluster File System. Research it, it's pretty good and flexible.

But no, sadly there is no pretty GUI-point-and-click-interface solution for it.

EDIT: Ask yourself though if you really need this HA storage clustering, or can you just live with nicely done backup system and RAID arrays with spare drives (if you feel extra vulnerable).

share|improve this answer
I have enough data to where if my RAID card fails or filesystem(s) get corrupted, I could be down for 1-4 weeks while I restore from a backup. That would be disastrous. – Phil May 21 '12 at 21:30
1  
Clustered filesystems might not help, if you corrupt or delete a large number of files and they replicate those changes are you still down for 1-4 weeks to restore? – EightBitTony May 21 '12 at 21:40
That's a good point. The main failure case I was thinking of is hardware failure. In the case of file system corruption (like MFT corruption), I would hope the cluster would be smart enough not to replicate those problems. – Phil May 21 '12 at 21:43
@EightBitTony That's why you have snapshots and lock your access to an api :) – Tacticus May 22 '12 at 1:44
up vote 0 down vote

If you're familiar with Windows I suggest taking a hard look at Microsoft Windows DFS. It offers file based replication between Windows servers, presented as a single share to the users.

But, it depends to some extent on what you want to present the storage to.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.