Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

chore: Optimize fetching samples logic #25995

Merged
merged 1 commit into from
Nov 16, 2023
Merged

chore: Optimize fetching samples logic #25995

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
23 changes: 10 additions & 13 deletions superset/views/datasource/utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -104,21 +104,18 @@ def get_samples( # pylint: disable=too-many-arguments,too-many-locals
result_type=ChartDataResultType.FULL,
force=force,
)
samples_results = samples_instance.get_payload()
count_star_results = count_star_instance.get_payload()

try:
sample_data = samples_results["queries"][0]
count_star_data = count_star_results["queries"][0]
failed_status = (
sample_data.get("status") == QueryStatus.FAILED
or count_star_data.get("status") == QueryStatus.FAILED
)
error_msg = sample_data.get("error") or count_star_data.get("error")
if failed_status and error_msg:
cache_key = sample_data.get("cache_key")
QueryCacheManager.delete(cache_key, region=CacheRegion.DATA)
raise DatasetSamplesFailedError(error_msg)
count_star_data = count_star_instance.get_payload()["queries"][0]

if count_star_data.get("status") == QueryStatus.FAILED:
raise DatasetSamplesFailedError(count_star_data.get("error"))

sample_data = samples_instance.get_payload()["queries"][0]

if sample_data.get("status") == QueryStatus.FAILED:
QueryCacheManager.delete(sample_data.get("cache_key"), CacheRegion.DATA)
Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm not entirely sure why we only delete the sample data cache key, however I thought it was best to preserve the existing logic.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The "count star" might be a heavy workload in some db, it trigger full table scan and aggregation, so just removed "sample_data" but keep "count star", it means that even though "sample_data" query is failed, count_star_data is cached still.

I think the new logic should remove cache key if the "count star" query is failed.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

and...another think is that we should customize count(*) in different db.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks @zhaoyongjie for the comment. Thinking about this more, if a query fails then there’s shouldn’t be anything cached. Hence, in the new sequential formulation, if the COUNT(*) query succeeds but the sample data query fails we should actually be removing the cached result from the former.

raise DatasetSamplesFailedError(sample_data.get("error"))

sample_data["page"] = page
sample_data["per_page"] = per_page
Expand Down