Google Cloud Bigtable: petabyte-scale NoSQL database for scalable computation
summary
Cloud Bigtable is Google’s petabyte-scale NoSQL store powering Search, Maps and Gmail and it exposes the same architecture via Cloud Bigtable with Colossus storage, HBase-compatible API, Go/Python clients and CBT. Clusters are zonal with SSD or HDD nodes, minimum three nodes and linear scaling roughly 10,000 QPS per node so 3-node minimum -> 30k QPS and customers run up to 3 million QPS at 300 nodes. Typical 99th‑percentile latency is ~6 ms and rows are limited to about 100 MB.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 0:35Agenda
- 7:00Cloud Bigtable
- 15:00Login to Cloud Console
- 16:00Getting started with Cloud Bigtable
- 20:35Quick example
- 21:52Operations
- 24:05Cloud Dataflow
related talks
05mNext Big Step
09mExplains the Cloud Bigtable Platform
33mManaging IoT Storage with Google's Cloud Platform
06mGoogle BigQuery for enterprise
40mDeveloping on Google Cloud with Python and DataFrames
01h 02mDiamond Sponsor Hall
52mHow to build a planet-scale data replication...
24mWhy Google Cloud is the Best Place for Your PostgreSQL Workload
