Common O&M issues and solutions for OpenSearch.
Large data volumes and inaccurate statistical results
-
Each searcher retrieves results, performs a rough sort on 1 million of them, and then performs a fine-grained sort on 200 of those results. After a final sort, it returns the top 5,000 results.
-
When the data volume is large, the system sorts a random sample of the data and estimates the final statistics based on this sample. As a result, the statistical values are not exact.
Processing datetime and timestamp data
-
OpenSearch automatically converts datetime and timestamp values in RDS tables to milliseconds.
-
To preserve the original time format, convert the data to a string in RDS before importing.
If the issue persists, contact technical support.