✳flâneur — a map of the web's best reading

https://pdos.csail.mit.edu/6.824/notes/l-spanner.txt

pdos.csail.mit.edu · 1,467 words · saved by 1 readers

6.5840 2023 Lecture 14: Spanner Why this paper (Google Spanner, OSDI 2012)? A rare example of wide-area distributed transactions. Very desirable. But two-phase commit viewed as too slow and prone to blocking. A rare example of wide-area synchronous replication. Neat ideas: Two-phase commit over Paxos. Synchronized time for fast r/o transactions. Used a lot inside Google. What was the motivating use case? Google F1 advertising database (Section 5.4). Previously sharded over many MySQL and BigTable DBs; awkward. Needed: Better (synchronous) replication. More flexible sharding. Cross-shard transactions. Workload is dominated by read-only transactions (Table 6). Strong consistency is required. External consistency / linearizability / serializability. The basic organization: Datacenter A: "clients" are web servers e.g. for gmail data is sharded over multiple servers: a-m n-z Datacenter B: has its own local clie

6.5840 2026 Lecture 12: Spanner Why this paper (Google Spanner, OSDI 2012)? Unusually ambitious for its time: Wide-area distributed transactions. Consistent wide-area replication. Neat ideas: Two-phase commit over Paxos. Read/only transactions w/o locking, via Snapshot Isolation. Read/only reads from local replica, but fresh. Synchronized time for external consistency. Spanner is used a lot inside Google. And influential, e.g. CockroachDB. What was the motivating use case? Google F1 advertising database (Section 5.4). Previously sharded over many MySQL and BigTable DBs; awkward. Needed: Synchr

Explore this link on the map →

related reading