RMOUG 12 — Rocky Mountain Training Days

Well, I’ve just returned from a fabulous trip to the Rocky Mountain Oracle User’s Group (RMOUG) training days conference in Denver, Colorado.

It’s been a long time since I’ve attended this conference, but in the fall when I started to plan out which conferences I would begin re-attending in 2012 I gave this one a solid look. Often it’s hard to choose among all of the good conferences, but I thought I’d limit myself to the top 4-5 conferences based on my interests, and this one made the cut.

The main reason I wanted to attend is that I’m targeting conferences that try to bring many groups together instead of being a giant echo chamber for a homogeneous group of people. Some regional Oracle conferences start to seem like DBA-clubs at which the same topics are re-hashed again and again (the basic ones talk about space management and backups, while the advanced ones talk about block dumps and index leaf node management).

When I looked at the agendas for past RMOUG conferences, I was pleasantly surprised to see a wider range of topics that included DBA topics AND developer topics — and it’s this intersection that has me so keenly interested lately. Much of what I’m seeing in the data world is increasing conflict between DBAs and developers that is resulting in poor architectural and process decisions all-around. I’m hoping that conferences like RMOUG can be a venue for facilitating discussions between the camps.

The conference itself was run like a clock — but in an interesting way. It’s the most laid-back well-organized event I’ve ever been to. You know how some conferences are scheduled to the hilt with barely enough time to sprint from session to session and event to event? RMOUG isn’t like that — there’s a quiet competence about how its run so that you don’t miss a thing, and you don’t feel rushed or stressed about it. If you know the people involved, you probably understand why — Tim Gorman, Kellyn Pot’vin, John and Peggy King all exemplify that quite competence and energy that makes sure stuff gets done without a big deal being made of it.

Anyway, I went to some great sessions by Kris Rice (on the Oracle Database Cloud), Graham Wood (on under-utilized Database features), George Thrower (on Ruby!), and Jean-Pierre Dijcks (on the Oracle Big Data Appliance). I also caught up with lots of my OakTable friends.

In Kris’ session I learned a lot about the differences between the Oracle Database Cloud and the Amazon Relational Database Service (RDS) — basically Amazon RDS presents you with an instance and the ability to connect just like any other instance through a machine and port, while the Oracle Database Cloud is more like a schema with more limited access and extreme security (the Oracle Cloud lacks the Amazon security controls around IP address access, so it settles for more constraints around access). Also, file and object deployment is heavily virus scanned on the Oracle Database Cloud, leading to multi-step deployments — Kris showed us using the Deployment Cart functionality of SQL Developer. Also, a lot of emphasis on using RESTful interfaces to the Cloud (with a quick side note on the ability of the APEX listener to expose database capabilities as RESTful interfaces WITHOUT requiring a full APEX installation!).

In Jean-Pierre’s presentation I learned a lot about the connectors from the Big Data Appliance to an Oracle database, including specialized Hadoop finalization jobs which take reduce output and format it specifically for insertion into an Oracle database, along with an optimized hdfs client that will enable using hdfs content as external tables. JP described tests that were able to move 15TB/hr from hdfs on a BDA into an Exadata machine over Infiniband.

My own presentation on writing MapReduce in SQL and PL/SQL was rather sparsely attended, but that’s ok — I understand that presentations on the edge between DBAs and Developers aren’t always the most popular. I think it went well and my audience got some good stuff out of it. It was my first run of this presentation and I’ve taken away some good ideas on how to improve it.

In any event, from Cary Millsap’s keynote (in which I loved the difference between memorizing facts and understanding concepts as a way to “master” material) to Kerry Osborne’s presentation on DIY Exadata, I had an excellent time and will be sure to try and attend next year.

Between the location (Denver and the mountains), venue (Colorado Convention Center), and the excellent content I’d highly recommend putting this one on your list.

I wish I could go inside and see the RMOUG sessions!