You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We currently annotate S3 GET requests as "Async" or "Sync" and pass this into the referrer header. This is useful when looking at access logs and performing regression analysis.
This annotation can be improved/fixed to make this analysis easier:
1/ All sequential prefetch requests are currently marked as SYNC, even though these are ASYNC.
2/ All parquet column prefetch requests are marked as ASYNC, we can make this more granular by adding more read modes: COLUMN_PREFETCH, FOOTER_PREFETCH, REMAINDER_COLUMN_PREFETCH
This will allow us to quickly identify where an S3 request came from.
Code of Conduct
I agree to follow this project's Code of Conduct
The text was updated successfully, but these errors were encountered:
Tell us more about this new feature.
We currently annotate S3 GET requests as "Async" or "Sync" and pass this into the referrer header. This is useful when looking at access logs and performing regression analysis.
This annotation can be improved/fixed to make this analysis easier:
1/ All sequential prefetch requests are currently marked as SYNC, even though these are ASYNC.
2/ All parquet column prefetch requests are marked as ASYNC, we can make this more granular by adding more read modes:
COLUMN_PREFETCH
,FOOTER_PREFETCH
,REMAINDER_COLUMN_PREFETCH
This will allow us to quickly identify where an S3 request came from.
Code of Conduct
The text was updated successfully, but these errors were encountered: