Monday, April 23, 2018

COLLABORATE18: Day 1

Yesterday you probably noticed a trend, and today I can tell you was NOT all about me. :} Instead of running or helping run SIG meetings, I decided to spend my day getting outside of my comfort zone and getting into sessions that were not in my wheelhouse so I would have to stretch and pick up new skills or at least knowledge/terminology!

Building A Continuous Development Pipeline for WebLogic DevOps - Eric Mader (@ericgmader) - 9:45 AM

Eric sure knows his stuff! We do have CD were I work, but we are not fully DevOps in my area...yet...AND I have been working on WebLogic for the past year so this was too good to pass up! I liked it, and it helped reinforce some things like the People, Process, Technology pyramid, that I have to read The Phoenix Project book, and what some obvious benefits of DevOps are. Oh, and it was interesting how there is a new role coming out called "Operation Developer" which really embraces DevOps and what a person has to do here in this space.

Introducing Kafka to the Oracle DBA - Mike Donovan (@dbvisitmike) - 11 AM

I was totally not ready for this session, and that is great! It was ALL completely new to me!! "The truth is the log. The database is a cache of a subset of the log." Whooooooaaaaaa!! Boiling Kafka down to the nitty-gritty, it is pretty much just a log writer/reader that is really good at what it does. I need to get over to http://confluent.io to see about getting some free books which may include Kafka: The Definitive Guide book, or otherwise just find it online at a retailer since this seems like a wide open space which is interesting.

The Self-Driving Oracle Database of the Future (or Present) - Rich Niemiec (@RichNiemiec) - 12:15 PM

DBAs need to become data experts, not database experts to survive the future, which includes moving closer to the business and innovation. Awesome stuff to hear! Data will become the new oil, which we see with Tesla becoming worth more than General Motors who shipped 92 TIMES as many cars as Tesla in a quarter recently. On-Prem 18c DB WILL exist. Also, try it out by going to http://cloud.oracle.com/tryit with some free credits.

Going Serverless - An Introduction to AWS Glue - Michael Rainey (@mRainey) - 1:15 PM

In this context, serverless is going to a fully managed provider which allows developers to focus on developing instead of supporting. There are challenges facing this such as: monitoring, debugging, integration testing is difficult, maintaining state of application across multiple functions is tricky, startup latency, built for small short functions, vendor lock-in. AWS Glue - "a fully managed extract, transform, and load (ETL) service" straight from their manual.

General Session: Oracle Cloud - How to Build Your Own Personalized Path to Cloud - Steve Daheb SVP (@OracleCloud) - 2:30 PM

This one was a lot of general things I have heard before, including some data/examples Rich shared at lunchtime (like AWS only offering 99.95% availability which comes out to 20 minutes of downtime a month NOT including several other reasons there could be an outage, yet Oracle is offering 99.995% which means less than 2.5 minutes of downtime a month WITH NO CAVEATS), but one thing I heard which caught my attention was Oracle Autonomous Cloud is coming at some point, in addition to the OLTP version of this (18c) coming this Summer.

Advance from DBA to Cloud Administrator - Erik Benner (@Erik_Benner) - 4:15 PM

This echoed a lot of what Rich was saying too, where DBAs need to go from Database Management to Systems Management. Example of how many people in the audience (most) that were using Outlook 365 which is just an instance of a workload/system moving out of your DC and into the Cloud. New skills you need to get include: DATA administrator, small environments (focus on the application), large environments (focus on all of that plus, automation, security, integration). Brand new product to help with this called Oracle Management Cloud, which is NOT replacing Oracle Enterprise Manager, but this is an evolution of OEM in that it is using a big data repository to put just about everything into the same bucket so it can learn about the whole system. After it learns, it allows you to drill into what it has aggregated so you can get down to the specific transaction/SQL statement/error message at any time in the past. Heady stuff!!

No comments:

Post a Comment