Explanation:
Bigtable is ideal for high-performance, scalable storage of structured time series data like CPU and memory usage from millions of computers. A narrow table design in Bigtable, with a row key combining the computer identifier and sample time each second, ensures efficient storage and retrieval for real-time analytics. This approach supports detailed analysis at the second level. Other options fall short: BigQuery isn't optimized for such granular time series data (A & B), and aggregating data by minute (D) limits analysis granularity and real-time querying capabilities.
You need to store time series data for CPU and memory usage from millions of computers, with data recorded every second. Analysts will perform real-time, ad hoc analytics on this data. The solution must avoid per-query charges and scale with the dataset. Which database and data model is best suited for this scenario?
A
Design a wide table in Bigtable, using a row key that merges the computer identifier with the sample time each minute, and include second-level data as columns.
B
Set up a table in BigQuery, continuously adding new CPU and memory usage samples to it.
C
Implement a narrow table in Bigtable, with a row key that combines the Computer Engine computer identifier and the sample time each second.
D
Construct a wide table in BigQuery, with a column for each second's sample value, updating the row with each second's interval.
No comments yet.