Information Playground

2008 Solar Stats

With over a foot of snow landing on my roof this past weekend I do believe my solar panels have generated their last watt of 2008. This was the first full year I had them so I thought I would tabulate their monthly output for the entire year.

 

I'm pretty happy with this data as it makes up roughly 20-25% of my usage. The downside of course is that there are 2-3 months out of the year where there's a good chance they'll be non-functional depending on snowfall totals.

Below is a cool picture that my contractor put together. He actually created this picture for my son, who organized a "Solar Energy" town meeting for his Eagle Scout project. My contractor was NexAmp.

Steve

Obama and Digitized Healthcare Records

I was reading a Wall Street journal article about President-elect Obama's "to-do list" for the five broad spending areas that his administration plans on pursuing. The fifth one caught my eye: Digitized Health Records.

A quote from the article:

"the  country's $2 trillion health care system is clogged with paper, folders and  plenty of clipboards. Some 90% of U.S. doctors and at least two-thirds  of hospitals still rely on paper patient records, and many of those who have  gone digital can't exchange the information with outside providers. The result  is billions of dollars each year in administrative waste, duplicate tests and  medical errors"

I'll be interested in watching these U.S. government decisions roll out.

Will the US make the same decision that Finland did?

Contest News

The flurry of activity from October's EMC Innovation Showcase seems like a distant memory, but I did receive an interesting email yesterday from Burt Kaliski (who works in EMC's office of the CTO):

Dear Steve --

As "Solving the 100 Year Problem" is a recipient of one of the top three awards, the CTO office will be taking the lead on incubating the idea, as Jeff Nick announced at the Innovation Conference.

Dictionary.com provides a good illustration of what "incubation" involves

> "to maintain at a favorable temperature and in other conditions promoting development"

The CTO office will stay with this idea through its natural development and adoption by the company, should the business case prove out.

An architect from the Advanced Technology Ventures team will be assigned to the idea. The CTO office may also assign advanced technology resources at the China CoE, and engage with business units, as appropriate to advance the concept.

Let me know if you have any questions, and congratulations again on your idea on reaching this point in the innovation lifecycle!

-- Burt

This is pretty cool as I had a good experience last year working with the team in China on the Data Lineage idea (which eventually was put on display for customers at EMC World last May).

Steve

Digital Curation Conference

I mentioned in a previous post that a paper I had written for the 4th Annual Digital Curation Conference in Scotland was accepted and presented in Edinburgh on December 3rd. The paper was about XAM.

I wasn't able to travel to present the paper but one of my co-workers from Centera presented in my place. I asked him how it went and he said roughly 30-40 people attended and that there were lots of heads nodding.

I do believe that by "heads nodding" he meant that the XAM message resonated with the Digital Curation audience, as I expected that it would.

I should have pressed him further to clarify his choice of words because sometimes terms take on different meanings between the US and the UK. Hopefully head nodding does not imply "dozing".

This post provides a link to the paper in case you wanted to nod your own head (for whatever reason).

Mechelen Closing

As reported yesterday, EMC's Centera development will be moving away from Belgium and the Mechelen development center will be closed. 

I certainly found this news to be disappointing, as the software developers in Belgium have always presented me with some great opportunities to collaborate and build interesting products. In fact we were just on the phone last week brainstorming about new directions. I have always found these discussions quite exciting. Best of luck to them.

I could make a very strong argument that the work done in Mechelen was some of the most disruptive and innovative storage software development of this decade.

In that spirit I'd like to highlight what I think is their main accomplishment.

They took a desktop application (Filepool) and turned it into a 5 9s storage system.

Expressable Widgets

I graduated from college with a pretty solid understanding of how to build software.

I was able to hold my own in many design meetings with other engineers, managers, and test folks.

But I was unprepared for certain meetings that I had with marketing.

Not every meeting, mind you. Just those particular meetings with marketing co-workers that were "overly-curious" about my software design.  I used to be caught off guard in these situations.

Not any more.

As a public service I thought I'd relate a technique that a mentor of mine here at EMC uses quite often in these situations. It's a method that's invaluable for junior developers.  I've seen it work, and it's powerful.

This technique is called expressable widgets.

To Manage or Not to Manage

I've been reading Spence's posts about becoming a manager with some interest, given that I hadn't heard the full story before. Last time I worked in the same building with Dave it was right after I suckered recruited presented him with the opportunity to build StorageScope after I had worked on the initial research and design.

I find it even more interesting because Dave is one of the better software engineers that I have worked with (we collaborated quite a bit on building Navisphere ). When Dave became a manager, I always wondered, why did he do that? Now I know!

His story about becoming a manager has me reflecting on why I never did. It's a good case study of two technical people, one who successfully moved into management, and one who never even tried. My reason is pretty simple.

I like building stuff too much.

NICE Phone Calls

When I think about the type of fixed content that is traditionally stored on a Centera there are several use cases that come to mind.

Medical images. Scanned check images. Mortgage applications. Anything that comes from a scanner. E-mail.  Electronic files.

All of these use cases represent content that is not likely to be "opened and edited".

Lately I've been reading a lot about another use case that makes perfect sense for Centera.

Phone calls.

Picture a large call center handling 1000s of calls each day. As soon as each call ends, the entire conversation becomes fixed content.  Months later a customer may file a complaint about a service that was promised during a particular phone call.  How can you find the call?   Was it deleted?  Is it authentic?  Is it on tape somewhere?

One of the worldwide leaders in multimedia recording solutions is NICE, and NICE has integrated Centera into their solution.  For example, NICE's Perform product archives phone calls to Centera and then provides a wide suite of analysis tools such as word spotting, emotion detection, and voice analysis. This type of capability is valuable in any market that utilizes a large call center, including public safety, finance, retail, etc.

Integration Details

NICE and Centera integrate via NICE's Storage Center Server (SCS), an architectural component which is responsible for storing and tracking all phone calls entering the system.  The SCS pulls previously stored calls from NiceLog Recorders (pictured below), stores each call to Centera, and receives a Centera content address (CA) in return. This CA is stored in an SCS database along with other information about the call.

The SCS contains software that has integrated with the Centera SDK. This SDK integration takes advantage of the following Centera features:

• Multi-pathing: The SCS administrator can enter the IP addresses of up to four Centera ports in order to realize load balancing and multi-pathing to the Centera system.
• Automated DR failover: The SCS administrator can also enter the IP address of a replica Centera system in order to continue in the face of a disconnect from the primary Centera.
• Metadata annotation: The SCS is able to use Centera's metadata feature to annotate the phone call being archived. For example, the name of the SCS application can be stored along with the phone call.
• Retention: The SCS always stores a retention field in the Centera metadata associated with each phone call. Centera enforces "disallow delete" based on this field. The time is based on the call arrival time (not call archive time).
• Multi-Threading: The SCS fires up multiple threads (default of 20) to allow parallelization in the ingest of phone calls. This is particularly useful if batches of phone calls are archived at night, for example.
• Shredding: Once phone calls do expire, there is often a need to shred the phone call electronically. Centera supports this.

The Centera integration has been successful to the point where NICE refers to a Centera deployment as a "best practice". Over 100 NICE/Centera deployments are in the field. The NICE/Centera integration has eliminated a lot of headaches that used to occur with phone call archiving, including finding/loading tapes, proof of originality/authenticity (provided by Centera's content address), and configuration-free capacity upgrade.

NICE Wholistic Approach

I don't want to shortchange NICE by highlighting just the audio aspects of the Perform product. NICE has a much larger portfolio that can also archive customer interactions such as email exchanges, web interactions, video, etc. NICE provides wholistic management of all of these use cases (which all represent fixed content, by the way).

What I did want to highlight was the unique value proposition gained by NICE several years ago due to the fact that the NICE developers integrated with Centera. 

One more interesting point of note: converting the NICE SCS to use the XAM API would not only maintain compatibility with already stored audio files, but it would also provide NICE customers with the flexibility of moving to future (XAM-based) archival systems.

Steve