Can’t win ’em all – darn it!
A year ago I offered predictions for 2007 (see 2007 in review) and it was with deep foreboding that I went back to look at them.
Cynic or optimist?
Here’s the list and my take on each prediction.
-CIOs realized they are at a significant cost disadvantage and started prototyping new apps on services such as Amazon’s EC and S3. They did see some cost savings, but the big win was faster implementation and scaling of the infrastructure. Very little industry buzz though, since storage vendors don’t do business with the service providers and the CIOs who know aren’t talking.
Cleverly worded to avoid measurement, I suspect that while some CIOs have piloted projects using online resources, the practice is not as common as I’d thought.
Medical records market
-Electronic medical records failed to get any storage industry attention even as this multi-billion dollar opportunity slipped away. American consumers, seeing widespread abuse by insurance companies, debt collectors, law enforcement and employers, began fighting their use in the courts and through “off the grid” health care options.
Nailed the industry’s reaction to EMR. Are American’s fighting their use? Hard to say. It appears they haven’t gotten enough traction to spur a reaction one way or the other.
1.5 TB disk drive?
I predicted they’d be shipping by year end, but none have even been announced. Is disk capacity growth slowing? We’ll know by mid-year.
Small-write optimized flash drives?
-The biggest device surprise – Small Write-Optimized Flash (SWOF) drives – in 32 and 64 GB capacities from Samsung, took off in as OLTP-oriented IDCs realized they could have higher reliability and performance at much lower costs over conventional disk drives. In a brilliant marketing move, Samsung got certified with Oracle, DB2, SQL Server and MySQL, dramatically raising the technical buzz around SWOF drives.
If we had more performance data Fusion-io drive might be a SWOF drive. And no Internet data centers picked them for broad implementation. Total miss.
Zmanda open source backup
-Zmanda open source backup really started taking off this year. It is cheap, reliable, and easily readable. At some point folks started waking up to the fact that they could lock themselves in to a single software product by using some sort-of cool features, or they could use Zmanda and know they’ll be able to recover the data on any Unix/Linux system.
Zmanda CEO Chander Kant assures me that Zmanda sales indeed started taking off this year. I’ll take his word for it.
He also noted that they can back up directly to Amazon’s S3. According to the AWS blog:
Zmanda Internet Backup is a plugin for the Amanda Enterprise backup software. Amanda Enterprise is a certified, tested, and supported version of the popular Amanda open source backup and recovery tool. Amanda can now use Amazon S3 to backup, archive and retrieve any amount of data, at any time, from anywhere on the Internet.
Apple’s Time Machine
-Apple’s spring trio of storage software products, based on ZFS, including Time Machine, Replicant and Total Recall – note their cool use of Sci-Fi movie references – left every other storage app looking like crank-start Model T’s next to a Ferrari F430.
No other cool storage tools. No cool names. Fact is, I’m not even using Time Machine yet because I’ve had so many problems with Leopard.
But I’m sure it will be insanely great when I do.
The final prediction:
2007 was a good year for storage. Despite the profitable inertia enjoyed by the big guys, some exciting technology and applications made waves . . . . All of these contributed to a growing rethinking of storage architectures in the high-volume, cool-data, internet age.
I stick by this last prediction: 2007 was a good year for storage. The momentum for cluster-based storage grew, particularly with EMC’s Hulk/Maui teaser announcement. 1 TB drives made it into mass production and their prices are dropping. WD’s greener drives appear to be a real advance.
The StorageMojo take
My predictions were generally optimistic so I guess I must be an optimist. But the big trend in new storage architectures is one I’ve been looking for – and advocating – for years. And that is definitely happening.
While I don’t buy the hype around cloud computing, it is clear that big customers and some VCs are waking up to the fact that the old storage business model is not going to be the new storage business model. We are on the verge of a surprisingly rapid shift in how new data systems are designed and built.
But that is a topic for another day.
Comments welcome, of course. What was the biggest surprise you saw in 2007?
Not bad track record, better than our economic advisers, the housing and job markets. Maybe you’d do better if you switched to virtual predictions ?