时间:2021-07-01 10:21:17 帮助过:3人阅读
This year I spoke about synchronous replication at a higher level. The talk was
titled“ACIDic Clusters: Review of current relational databases with synchronous replication”. Having previously given talks with boring titles (but interesting content), this time I decided to go with an interesting title, and it seemed to fit well with topic being discussed.
In short, the talk was about ACID-compliant databases supporting relational
semantics (essentially, SQL) while providing synchronous replication. The talk
was not about why we need all three in a database, but it was on why some of
them may depend on each other. In this review, I looked at various aspects and
features of databases like Galera/PXC, MySQL NDB, Google F1 and FoundationDB,
while also discussing the topic of synchronous replication and its family in
general.
One of my primary motivations for this talk was so that people truly understand
what they are using, specifically in my case, the product that I manage/develop
- Percona XtraDB Cluster (PXC), based on Galera replication. There are plenty
of misconceptions on what a synchronous replication needs to be, and what it
doesn’t, and on the overheads/latencies associated with them. It was also
interesting learning how other systems like F1/Spanner are built, and how they
handle latencies for instance or have constraints like atomic clocks. The issues
associated with transactions involving optimistic concurrency control than
pessismistic locking are also interesting to look at.
Here are the slides of the talk: