"The hows and whys of a distributed SQL database" by Alex Robinson

[@michaeleaster1815]

This is an outstanding talk. I appreciate the historical context/comparison of the architectural approaches.

[@vimalma1093]

Must watch for anyone trying understand distributed databases. Great talk and one of the best.

[@sidharthanrajendran1387]

Thanks for giving an comprehensive information about db from 1960s to now.

[@n8allan]

Great talk; I appreciated the technical details, while remaining high level.
I suggest using "database systems" when talking about systems, and "databases" when talking about an instance.
I also think it's safe to describe NoSQL systems as equivalent to the storage layer of SQL systems. NoSQL = distributed storage system
SQL = storage system plus query processor
NewSQL = distributed storage system plus query processor

[@stianmaurstad]

Very interesting,
thanks Alex for a great talk & thanks for sharing strangeloop.

[@Kserx.p]

very nice presentation with lots of detailed information and the rational of the distributed datastores

[@benevolent6705]

I agree with previous comments, this talk was very concise and informative.

[@sherryx4264]

great talk! it includes all the pros and cons of all types of databases.

[@gmcodes3797]

I hate cockroaches 😂. More seriously this was a pretty insightful talk, I have worked across many layers of the stack data, compute, frontend, orchestration etc. But databases are like the foundations of a building. Always super interesting.

[@karlpokus]

Great talk! Thanks for sharing.

[@nikolay6700]

Can DB use both: MVCC and locks. MVCC for reading and locks for writing?

[@yuhangmiao2796]

Yes. Spanner uses MVCC for read-only transactions and 2PL for write.

[@arunkutube]

awesome talk