DBA, Instrumentation Evangelist, (Grand)Father, Fosterer, Husband. Senior Director of Development for MySQL HeatWave Service (MDS) @ Oracle (views here my own)
@samlambert Insights on all platforms is built upon the shoulders of MySQL Enterprise Monitor, and it pushing query insights data in to performance schema.
Pour one out for MEM. 🍻
Last day at #CloudWorld. Find out how to benefit from integrated and automated machine learning using HeatWave AutoML and learn how MercadoLibre efficiently operates over 7,000 MySQL instances in the cloud.
https://t.co/CxoWlQwfdn
Want to deep dive into the latest HeatWave and MySQL innovations, hear from customers using them, and learn what’s coming next? Be sure to attend Nipun Agarwal’s sessions at #CloudWorld. Add them to your schedule now:
https://t.co/OwtqYMXKeq
This famous video filmed at the Sasquatch Music Festival 2009 is often shown in corporations, team working classes, etc. and it actually shows an interesting aspect in psychology: the role of leading with a example (even if weird) and imitation.
Join Edward Screven and Nipun Agarwal to learn how Oracle HeatWave enables you to take advantage of #GenAI—no AI expertise required!
https://t.co/5rsothyOYT
@sunbains@tzkb Up until MySQL 3.23 there was only the ISAM engine, no transactional engines. In 3.23 the engine interface was added to include BDB (later removed) for a transactional engine, with MyISAM added as the new default from an extended ISAM. All other engines extended from there...
The idea of locks being the only mechanism for isolation to guarantee the consistency are long gone. When these ideas were being debated, the data sets were very small, compared to today and distributed meant another computer in the same building somewhere (mostly).
The state of the art in transaction coordination and consistency mechanisms has gone beyond distributed locking because it just doesn't scale. It has shifted in favor of algorithms that use time based coordination in distributed databases.
@gwenshap My point is - whatever the language, whatever the app - if the task involved requires a lot of interaction with the data, then you want to run your code close to the data. PL/SQL can be a drop-in fix to lots of problems created where people didn't understand this axiom.
2/2