BlueArc and Drobo founder Geoff Barrall has a new perch: Gridstore, one of the companies I’ve been following for almost 3 years. Geoff is the new executive chairman. Formal announcement is expected this week.
Gridstore’s concept is a low-cost scale-out NAS appliance designed for office environments. Each box is a small, low-power node with a couple of TB. Stack ’em for as much redundancy, capacity and performance you want.
Think of it as the consumerization of hyper-scale technology. Nutanix writ small.
Gridstore details
Gridstore is offering a low-cost, scale-out network file server for $500 a node. That is too cheap for the enterprise storage companies to sell directly.
Founded 5 years ago, Gridstore got a beta out in 2010, and have been shipping for well over a year. They are a Microsoft CIFS protocol file server, using Microsoft’s storage server software. Running on small, 25 watt Atom-based boxes, a 6 node configuration is the size of a bread box.

Like other scale-out NAS systems, the Gridstore NAS has no single point of failure and can survive multiple node failures without going down or losing data.
They call their redundancy scheme RAIDg. When you set up a volume you dial in how many faults you want to survive and the software handles the rest.
Today the number of faults they can handle is limited to half the number of nodes minus one. If you have a 6 node configuration it can handle the loss of 2 nodes. They expect to relax that requirement in the future.
The StorageMojo take
Haven’t spoken to Geoff about this, but Gridstore seems like a natural for him. If there’s a theme to his many endeavors, its making advanced NAS technology more accessible.
Gridstore fits the bill nicely. If there’s one complaint about Drobo, its the lack of box-level redundancy. Gridstore answers this objection, at a higher price point.
Drobo – over 200,000 units sold – has blazed a trail for bringing advanced storage technology to the masses at affordable prices. They may be the first, but as Gridstore and others demonstrate, they won’t be the last.
Courteous comments welcome, of course. Hoping to make it to CES later this week. Readers: anyone I should make a point to see?
I love the idea of gridstore. My only beef with it, is it requires a proprietary driver on all clients accessing the volume(s).
That’s not the end of the world, but it killed the uses I was planning to make of it… (cheep storage for desktop folks to use while troubleshooting other computers). My windows team wasn’t willing to push the driver out everywhere.
This is definitely targeted to the smaller end of the SMB market. I consider us to be in the Medium portion of the spectrum (500 employees and 200 desktops), and the idea of pushing out and supporting software to every desktop just for file server access is a bit much. Especially when file shares on other NAS units are much simpler.
But the idea of off-loading storage calculations (stripes, assembly, etc) to the client (where there is usually a significant surplus of CPU) is VERY interesting. Much like EMC’s Avamar backup clients, it helps increase performance by leveraging what’s already there. Very interesting…
If I had to make any guesses, I’d think that eventually gridstore would want to produce some kind of node that could act as a CIFS (or other protocol) proxy (if not have the regular ones do this job on their own)
This is interesting. But isn’t this system prone to the same problem as Isilon, namely an abysmal power-to-storage density? Even with very low-power servers, you still have one server per drive or pair of drives. This can quickly become expensive to run.
Unlike Isilon, Gridstore is not aiming at the data center, where scale can quickly become an issue. Your average 5-35 person small business is unlikely to need more than a few TB of protected storage. At that level, the power costs are a rounding error compared to management and maintenance costs.
Could Gridstore pickup the low-end of the non-consumer NAS market where NetApp StoreVault made a foray into but quickly abandoned in circa 2008?
I suspect that is their intention. Beyond pricing though, this is really a matter of channel strategy: a direct sales force can’t sell $5-$50k products economically. The vendor has to either go through the channel, reaching the VARs and MSPs that service the SMB market, or go direct, as Drobo did early on.
StoreVault had another issue: too much functionality for the price. The direct sales force rebelled against a product with 80% of the functionality at 20% of the price.
I hope you like your direct sales person, because you’re paying for him.
Thanks for the interest in gridstore and the positive comments.
To clarify a couple of points, the virtual controller (driver referred to above) can be deployed automatically through a combination of active directory (for access control to volumes) and delivered from a domain controller as an msi file that is loaded when a user logs on (same as you would map volumes to users/groups/machines).
While the virtual controller technology gives us powerful advantages in both cost and performance – we also recognised from the start this can be a hurdle for larger organisations. We have invested a lot to simply and automate this process – and we will continue to advance this to where it.
Considering we have made scale-out storage available to the smb and mid-market without the cost and complexity of building clustered file systems, tweaking backplane networks and the waste of 3-way replicas common to most scale-out models – we believe distributing a virtual controller is a small compromise and one that we will work hard to automate so this is not a concern for even the largest enterprise.
The density of nodes will also be addressed shortly with 8TB nodes on the way. Most initial implementations are in the 10-20TB range with some customers already growing their gridstore to over 20 nodes while seeing io scale linearly.
Our current customers are looking to scale capacity and many could grow into the 100TB range or more. We plan to support our customers desire to scale larger gridstores while continuing to eliminate the need for complex clusters, backplane networks etc and offering customers simple scaling of capacity and io as they require.
Clustered models require high performance backplane networks to replicate data and can waste up to 50% of their IOPS on these background tasks. 3-way replicas are commonly used in clustered scale-out consume 3x capacity. Our customers who needs a 100TB does not want to buy 300TB and then have to house, power, cool and manage it.
With gridstore, data is encoded at source and uniformly striped across nodes in parallel and you get to use the full 100% IOPS. You get the protection against multiple node failures with Overhead similar to RAID.
Gridstore has boiled down storage to a simple set of building blocks that allows you to easily scale to any size you need – when you need it.
Gridstore product management.
Three issues right now during investigating, but not deal-killers yet.
One is the one-year hardware warranty compared against WD’s external drives’ three-year warranty (I know, not exactly the same thing, but it is what we are using for backups.) This may just be a size-of-company factor, that will change later.
The second is that I cannot find the product in searches on the websites of the resellers that we normally use, and there doesn’t seem to be a list of who is selling it on gridstore’s site. That’s not comforting right now – I’m sure it is just because of the “new-ness”, but still…
The last is nitpicky I guess, but I don’t see redundant power as a spec or an option – maybe it is there and I missed it. Unfortunately, in the a-little-too-big-to-be-small-world, redundant power supplies can translate to much needed reliability. (Or maybe I have just had horrible luck with power.)
Just my observations/ramblings.