Delta: Building Merge on Read

  Рет қаралды 1,884

Databricks

Databricks

Күн бұрын

Пікірлер: 2
@sreeramgarlapati9024
@sreeramgarlapati9024 3 жыл бұрын
Thanks for sharing your experience. Delta lake only comes with Copy on Write semantics. so, the idea of **using a combination of 2 tables & a view** - to build your own MOR semantics on deltalake sounded cool. Did you folks run into any consistency issues - as DeltaLake doesn't support multi-table transactions (docs.databricks.com/delta/delta-faq.html#does-delta-lake-support-multi-table-transactions) ? Meaning - when you are taking the change set and writing the changes to your baseTable - you will need a transaction b/w the base and changeSet tables... I was hoping if you folks had a general purpose solution - but this seemed to solve some very specific scenario. for ex: I couldn't find info. around - how are events replayed into the base table IN ORDER they arrived. for ex: cases like - if a single row (which can be identified by a unique id) - has been changed 2 times - & in the first change 2 columns were changed and in the 2nd change one of the columns in the previous change and few other columns are changed - how do you guarantee a safe merge. etc.
Intro to Databricks Lakehouse Platform Architecture and Security
28:47
Advancing Spark - Delta Deletion Vectors
17:02
Advancing Analytics
Рет қаралды 3,6 М.
One day.. 🙌
00:33
Celine Dept
Рет қаралды 65 МЛН
Delta Lake: Optimizing Merge
23:33
Databricks
Рет қаралды 14 М.
Delta Live Tables A to Z: Best Practices for Modern Data Pipelines
1:27:52
Eliminating Shuffles in Delete Update, and Merge
32:01
Databricks
Рет қаралды 4,9 М.
How Prometheus Monitoring works | Prometheus Architecture explained
21:31
TechWorld with Nana
Рет қаралды 1 МЛН
Advancing Spark - Understanding Low Shuffle Merge
18:51
Advancing Analytics
Рет қаралды 5 М.
Building Production RAG Over Complex Documents
1:22:18
Databricks
Рет қаралды 13 М.
[Webinar] LLMs for Evaluating LLMs
49:07
Arthur
Рет қаралды 11 М.
AI-Accelerated Delta Tables: Faster, Easier, Cheaper
39:13
Databricks
Рет қаралды 1,7 М.
3. Apache Kafka Fundamentals | Apache Kafka Fundamentals
24:14
Confluent
Рет қаралды 488 М.