The cost of MongoDB ACID transactions in theory and practice
summary
Covers MongoDB 4.0’s full ACID transactions and the related durability/consistency/isolation features, detailing available readConcern and writeConcern options and how they are implemented. Explains how those implementations affect client-side latency and compares the theoretical latency costs with measured real-world performance.
AI-written summary · about
chapters (12)
- 0:00Start
- 3:44Write Concerns Durability
- 14:10Simple update single threaded
- 18:41Multi-Az
- 20:10Multi-region)
- 22:10Atlas Latency Cakulator
- 25:33Consistency Levels & MongoDB
- 30:50Consistency Levels Details
- 32:12Consistency levels (1)
- 36:36Consistency levels (2)
- 38:03Consistency levels (3)
- 39:20Summary of main findings
related talks
01mDon't Drop ACID - Transactions in Distributed NoSQL
58mDon’t Drop ACID - Transactions in Distributed NoSQL
47mBeing transactional in 2019 - what does this actually mean?
52mTransactions: Myths, Surprises and Opportunities
52mACID in SQL Server: what it is, how it works, and how to live more adventurously.
01h 08mTransactions and Scalability in Cloud Databases—Can’t We Have Both?- 52mACID in SQL Server: what it is, how it works, and how to live more adventurously.
01h 00mDistributed transactions implementation trade-offs