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
Occasionally a dockless ETL task fails (sometimes due to network/HTTP/server errors, sometimes due to others), leaving temp tables (cf. #68) hanging around.
We should automate the removal of these. I see two strategies:
Upstream a patch to mds-provider that auto-drops them even if the ingest fails. There could be an argument for keeping them around for forsenic reasons, but I likely wouldn't trust them much in practice.
Make a new dag that automates dropping them, running perhaps monthly or weekly.
I'd lean towards option one, though they are not really mutually exclusive.
The text was updated successfully, but these errors were encountered:
Occasionally a dockless ETL task fails (sometimes due to network/HTTP/server errors, sometimes due to others), leaving temp tables (cf. #68) hanging around.
We should automate the removal of these. I see two strategies:
mds-provider
that auto-drops them even if the ingest fails. There could be an argument for keeping them around for forsenic reasons, but I likely wouldn't trust them much in practice.I'd lean towards option one, though they are not really mutually exclusive.
The text was updated successfully, but these errors were encountered: