Database Reference
In-Depth Information
Table 2.1
A Comparison between BigQuery, Cloud SQL, and Cloud
Datastore
Feature
BigQuery
Cloud
Datastore
Cloud
SQL
Data scale
Unlimited
Unlimited
< 10 GB
Supports ad-hoc queries
Yes
No
Yes
Supports fast data lookups
No
Yes
Yes
Can be used as a Hadoop source
or sink
Yes
No
No
Replicated across geographical
areas
Yes
Yes
Yes
Choice of datacenter locations
(EU and US)
No
No
Yes
Supports row-level updates
No (Append-
only)
Yes
Yes
Built-in historical snapshots
7 days of
history
No
No
Query UI
Yes
Yes (but not full
SQL)
No
Visualization via Tableau
Yes
No
Yes
BigQuery Service History
BigQuery releases a new update every week. Usually, the changes are small:
bug fixes, minor features, and other incremental changes. The rationale
behind this frequent release cycle is that it allows problems to be caught
early, allows major features to be phased in, and gives trusted testers the
ability to try out new functionality before it is released to the public.
This constant stream of updates is one advantage of a managed service
like BigQuery versus running the analytics software yourself; bug fixes get
applied automatically, and the system gets faster and more fully featured
over time. In addition, Google constantly upgrades its hardware and
improves its infrastructure components; this will translate into faster
queries and larger scale.
