Five Ways that NoSQL is like Fried Turkey
With the holidays among us, I must confess that I am a lover of the Fried Bird. It all started 5+ years ago when I was meeting with some guys on Tuesday morning and one of them told a hilarious story about his son-in-law nearly burning the house down by putting an empty aluminum pot on the fire (he wanted to dry it out before he filled it with peanut oil) and he forgot about it – being that it was Christmas Day, my friend did not get any fried turkey that year. :-( However, hearing this story made me curious about frying a bird and subsequently the next year I went to my local redneck sporting goods store and bought a turkey fryer – best investment I ever made.
Now, at both Thanksgiving and Christmas, I get up early in the morning and start the process – prepping the birds, injecting them with marinade, coating them with a nice spice rub, and preparing my 3 gallons of peanut oil (yes that is not a misprint, 3 gallons of peanut oil is used). At 3.5 minutes per pound at 325 degrees, what comes out of the oil is a crispy skin, tender and moist on the inside, most flavorful turkey you will ever eat! This year I fried 5 – that’s because my in laws all want to take one home for leftovers J – it doesn’t get any better than that for a cook. And for those that are concerned about all that oil being soaked up by the bird, even after frying 5 turkeys, I had the same amount of oil in the pot – these birds aren’t greasy.
Clearly fried turkey is a paradigm shift in the world of Thanksgiving and Christmas. Some might say the Turduken also falls into that category. And since we are all in the holiday spirit, I started thinking that just as the fried turkey is a shift away from the traditional roasting of the bird, so is Enterprise NoSQL a shift away from the traditional relational database platforms that we’ve all grown up with.
So, here are my 5 comparisons as to how NoSQL is like Fried Turkey
1. Paradigm Shifts – Both are relatively new to the marketplace. The relational database has been around for nearly 40 years, personally I grew up eating roasted Turkey for our holiday meals, it’s only in the last decade or so have we seen organizations starting to leverage NoSQL and people starting to eat fried turkey.
2. One size does not fit all – Although I love fried turkey, I’m also a big fan of stuffing, and unfortunately I have not figured out a way to stuff a bird with cornbread dressing and drop it into a vat of peanut oil without making a huge mess (No I’ve never actually done that – I’m not that stupid). :-) So my mother in law will still roast a turkey so that she can make her amazing dressing. Same goes for NoSQL, its not about to replace the relational database, but it can complement an organization’s existing infrastructure to help consolidate data from disparate silos of information – those being stored in relational databases as well as text, graphics, or other semi structured data.
3. Reimagining the taste of Turkey – I don’t know about you but I did not grow up with a love for turkey. In most cases I found it to be dry, the skin was soggy, and overall I would douse the bird in gravy just to get it down. Organizations that are using NoSQL are reimagining what they can now do with their data that they never thought possible before. Being able to ingest data natively without a lot of heavy ETL work, not having to build out complex schemas, and being able to use full text search on their data has given them a new understanding of their data because now they can get access to it like never before. They have a new love for the same data!
4. Speed – As noted above, frying a bird takes 3.5 minutes per pound. For a normal pot with 3 gallons of peanut oil, you don’t want to do more than a 15 pound bird (just search on you tube for videos of people not heeding this advice – not pretty) – thus within an hour you have your bird ready to go. Last time I checked it takes about 3 hours to roast a bird in the oven. MarkLogic is a schema agnostic NoSQL database, integrated with a search engine and an application server. These attributes add up to speed of application deployment. Imagine not having to worry about schema development, heavy ETL, normalization, as well as various tool integrations – MarkLogic customers are standing up applications 75% faster than using traditional methods. Why wait for the bird to come out of the oven, when you can have something more flavorful and crispy in under an hour.
5. Flavor – Injecting flavor into a turkey is a lot like giving your data context or meaning. MarkLogic is an RDF Triple Store which allows you to build relationships within your datasets within your organization as well as get really creative and start pulling in linked data from the web – i.e. DBpedia, etc.. Using the Semantic Web is the next evolution in data and can help further drive understanding and utilization of your data to ultimately impact your organizations top line revenue and bottom line costs. Next year I plan on branching out and mixing up various marinades. :-)
I hope you enjoyed my holiday blog and that you have a wonderful holiday season and happy new year!