- TS Queries
- Time Type Operations (0:00 ~ 3:00) Explanation is good. Please cut until this, and add explanation for to_timestamp_ms and to_epoch_ms.
- Selection and Interpolation -> please go through all available functions briefly. Here is the list, which should consists of time_next, time_next_only, time_prev, time_prev_only, time_interpolated and time_sampling.
- Add Aggregation operation, time_avg
- Use the kaggle environmental data when performing the time series queries. Using gs_sh is enough (prepare the command, and just copy and paste, while explaining what it does).
- For the population program, it is ok to add it as a bonus at thend if the users want to go over that blog, but prior to that all the time series functions should be ran and explained briefly via the gs_sh.
- Please do not forget to adjust the text size.
- Expiration Function/Data Retention
- I think only a few users would know that we had the row expiry release, so let's avoid saying that it is not supported in Java NoSQL interface or that the hash partitioning does not support it, and pretend that by default, interval partitioning is the best way to set the expiry release in GridDB because the nature of time series data.
- Use fixed list
- Based on the environmental data (from chapter 3) schema, create a new table (add new device) with the smallest retention period (3 days). Add new data in for the next 3 days and see on day 4 if any of the data has been deleted. Or maybe you can change the system date manually?
- Indexing/Partitioning
- Explanation is good. Let's add a simple instruction on how to add index , in SQL and composite index.
- (Comment: 8/30) Refer to Programming Guide for Advanced operations
Final Video for Chapter 4
Comment 9/21
Final Video for Chapter 4
Comment 9/21