This repository has been archived by the owner on Jan 5, 2023. It is now read-only.
SQLAlchemy connection pool overflow and TimeoutError on catalog-next #513
Labels
bug
Something isn't working
In doing load testing related to #449, the application was not responding correctly. A number of errors were encountered around serving standard traffic (
/dataset
,/organization
, etc). New Relic implied that the issue was in SOLR, but the SOLR instances were not showing extreme load and the configuration seemed appropriate.We found the following bug in the gunicorn log, and seems to imply an issue with sqlalchemy and/or the DB:
How to reproduce
Expected behavior
No overflow logs, traffic served is stable
Actual behavior
Overflow logs, traffic served is unstable/broken
The text was updated successfully, but these errors were encountered: