[FEATURE] Add id_expression to S3 integration MVs #1266
Labels
enhancement
New feature or request
good first issue
Good for newcomers
integrations
Used to denote items related to the Integrations project
Is your feature request related to a problem?
There's an incoming feature from opensearch-spark that will make including an id_expression mandatory for new materialized views. We don't have this currently, which will make future versions break if the issue isn't fixed.
What solution would you like?
Add an
id_expression
to each MV in integrations. It's likely thatuuid()
is sufficient, since the primary use case is log data where duplicates may occur.What alternatives have you considered?
Timestamps were considered as an ID field as well, but it's well known that even precise timestamps are not unique. For some integrations it may be possible to identify a log entry ID already, but I'm not sure if this is available for any of the current ones.
Do you have any additional context?
Necessary to unblock opensearch-project/opensearch-spark#143.
The text was updated successfully, but these errors were encountered: