I wish I created this video as it is engaging AND SUCCINCT. You might think 20 minutes isn’t succinct, but wow… an INCREDIBLE STORY is told in those 20 minutes and I highly recommend this video for anyone new to Trino to get you started as well as those that have been in this space a while to help you with your own storytelling.
I love that the creator walks through the whole reason we coupled data in systems like Hadoop. Data locality really did matter when the network became the bottleneck. I still say that the “no-network” model of moving the compute to data instead of the data to the computer will always be faster, but today’s separation of storage & compute offers many more benefits (especially cost & flexibility) that is is a the current right approach to data analytics at scale.
I also loved that the video creator talked about resiliency vs performance and even compared Trino with Spark in addition to Hive. Check out my hive, trino & spark features (their journeys to sql, performance & durability) post about how fault-tolerant execution is possible with Trino.
I can’t recommend the YouTube video above enough. Check it out!