October

Q3 is here and that's probably a good time to look back on how our industry did thru Q2 this year.
Here's the IDC numbers as commented by Computerworld and The Register: "Spending on all disk storage rose to nearly US$8.1 billion in the quarter, led by double-digit growth in emerging countries, IDC said. Enterprises also bought more bytes of data capacity, with shipments rising 24.8 percent from a year earlier to 6,667 petabytes"
So let's wait a couple weeks for these guys to add up the Q3 numbers for us!

On the networking side, I noticed a steady increase of articles and discussions about VXLAN so took the time to dig a bit into this. The one piece which probably provides the best introduction and overview about this networking overlay standard can be found here.
And while we are on the hype-side of the industry: immediately after we get used to SDN (software defined networking), the SDD (software defined datacenter) is upon us! I have yet to sort this out, seems to be a re-packaging of some sort of cloud: "The software-defined data center is a way for IT shops to reap the agility benefits of cloud computing while maintaining legacy applications."

Remaining on the topic of networking: If you are old enough to remember the SAN craze of the late nineties and early 2000, there was one company which stood out: Brocade! And one CEO who made a couple of millions and  headlines during these years as well: Greg Reyes. Well, a lot happened since then and you can now read his story in that book released just recently!

And finally: IBM just completed the acquisition of Texas Memory Systems (TMS) yesterday and their stock price is at an all-time-high on that same date... but if you think that Wall Street suddenly appreciates solid state storage, you're probably wrong!
By the way: IBM is already actively working on the next generation of "solid-state memory" as this press release from June indicates!

September

Looking out of the window,  (well, not my window!) seems like summer has come to an unexpected and early end here in Zurich. So the warm evenings are over, school is on again, traffic is heavy and it's back to business as usual for everybody...

Here's a couple news that drew my attention during the past couple weeks:

There's rumblings at Brocade: Their SVP WW Sales Ian Whiting left in June as it seems and CEO Mike Klayko announced his planned departure as well. Plus there seems to be reorganizations going on inside the company on all levels. Hard to say what's next for Brocade. For the time being, they still have a very solid lead in the legacy Fibre Channel market and recently seem to have a renewed focus on that part of their business (after being distracted too long with their Foundry-based Ethernet portfolio)!

IBM announced their plans to make it's next acquisition in Storage: Texas Memory Systems (TMS) is based in Houston, TX (as the name implies!) and has been in the solid-state storage market long before the industry invented flash memory! The acquisition is expected to close in the fourth quarter 2012, see details here.

Amazon made some waves with frozen water this month: Glacier is the (code?) name for their most recent cloud offering: This is a new type of storage cloud - aimed at the archive market and not inteded to provide split-second response times and data retrieval! Amazon claims to be able to offer this service at 10$ per TB per month. The jury is still out if they use tape in the backend or not...

And lastly some learning topics: SMB 3.0 (aka SMB 2.2) is being finalized and will be part of Windows 8 and Microsoft Server 2012. Looks like this protocol has the potential to definitely tilt the preferred Windows attach method from block to file!
And -together with Windows OS based services like "Storage Spaces"- make some HW based technology (like RAID) obsolete along the way! So watch out for this!

August

Very accurate positioning of Big Data and Cloud, two of the most beloved buzzwords of our industry today!
Couple additional comments: related to the "creation of information" I sometimes wonder where all these numbers come from and if taking blurry snapshots at a Saturday night party to be posted on Facebook really constitutes an act of "creating information"?!

To quote Jon Toigo: "I know of no empirical research that demonstrates with any validity the rate of data growth in a business. That’s because no one knows how fast their data is growing; they only know how much storage they’re buying year over year. The latter isn’t an indication of data growth rates, but of data mismanagement, pure and simple."

The New York Times just last week had an article about where and when Big Data started off, read it here.

Big Data involves analytics and data retrieval as we all know, and here's a report about recent IBM research related to that subject. SSD (solid state disk) plays an important role in the concepts and architectures to quickly retrieve and process data in a near real-time fashion.
Today, SSDs are still a rare sight in enterprise systems but this might change, once the industry is able to lower the price by reaching massive volumes.
And obviously, massive volumes can only be driven by consumer demand (like iPads and Laptops). This may be a couple more years away as this CW report and the recent Western Digital HDD sales numbers indicate!

Nevertheless, the SNIA has recently established a TWG (technical work group) to assess and define the impact of SSD to storage system architectures. Read here what Netapp has to say about this.

And lastly a listing of a few more recent articles, trends and events worthwhile noticing:

Leadership shuffle between VMware and EMC
LSI and Xyratex: 12 Gbps SAS is arriving!
Oracle: New tape library
IBM: How to tie PureSystems into your existing network
HDS: New Book: Storage Concepts: Storing and Managing Digital Data

July

Great reminder for those of you who did not grow up with mainframes: Here's a look back at storage tiering as it as was implemented on IBM Mainframes in 1989 (eighty-nine)!
Smarter Systems was called System-managed Storage back then (DFSMS) and did pretty much everything (and more!) clients expect from their storage systems today.
And it's still available on System Z, currently in its 12th release (V1R12.0)!

RoHS: you might have wondered what this is all about? RoHS stands for "Restriction of use of Hazardous Substances" and is a directive first issued by the EU in 2003. The RoHS directive requires that six hazardous substances be removed from all electrical and electronic equipment. The substances may be present incidentally at certain levels as long as they are declared.
On January 1st, 2013, RoHS II will become effective. Read here for details on what this implies.

And lastly some very exciting news again from IBM Research: The 12-atom storage!

June III

Raving comments on tape and why it is still very relevant at age 60!
If two posts within two days constitute a trend, then we have one here:

a) "Does EMC still think tape sucks? Nah, that is so last year – at least, if the storage giant's tech conference at the beginning of this month is anything to go by..."

b) " I know, tape was supposed to be dead 13 years ago. Gartner was paid a lot of money to say so in 1999 – in a report they now say they can not remember publishing. That report, which I read and complained about vociferously at the time, stated that 1 in 10 tapes fail on restore. Gartner now disavows this statement too in a classic Romney-esque bit of etch-a-sketchery."

And one major reason for all this excitement: LTFS and all the great opportunities it opens up. Lots off opportunities for the industry to be productizing that standard!

June II

The IDC numbers for Q1 2012 are out, here's the summary: "For the quarter, the total disk storage systems market posted just under $8.0 billion in revenues, representing 6.8% growth from the prior year's first quarter. Total disk storage systems capacity shipped reached 6,037 petabytes, growing 20.8% year over year."


Read the comments from "Computerworld" and "The Register".

June I

IBM's "Smarter Storage" announcements today!

According to this article in CW, "the National Center for Supercomputing Applications is rolling out a storage infrastructure that will include 380 petabytes of magnetic tape capacity and 25 petabytes of online disk storage made up by 17,000 SATA drives."

I think this is noteworthy for two reasons:
a) they realize that they cannot afford to keep all the data on spinning disk, so they provision about 95% of the data storage using IBM TS1140 tape drives and media (has someone invented the term "Tape Provisioning" already?)
b) the storage will be connected to the servers using 40 Gbps Ethernet, not sure if they actually use FCoE or something else (like AoE: ATA over Ethernet), but it still shows a trend towards high-speed converged networks. Ethernet has a great potential there with its 40Gbps and 100Gbps standards and roadmap!
In that project, they use the Lustre file system as the above article lines out. But they might as well have looked into using GPFS: IBM just announced the Rel. 3.5 with many enhancements as outlined here.
As you probably know, IBM uses the GPFS file system in many of its current storage offerings, like the TS7700, SONAS and most recently the Storwize V7000 Unified. So we can expect to see the functions outlined in GPFS 3.5 implemented in these products as well! Watch this space!

And to conclude todays excursion into the Big Data space, here's an article  from Jon Toigo which outlines quite nicely, what the difference between "lots of data" and "Big Data" is: “Big Data analytics help you make sense of what you are observing while you are observing it. You glean knowledge from data as quickly as it arrives — in something like 200 milliseconds – rather than waiting to batch process the data and produce a report.”