Getting thousands of TimeoutErrors errors

We used Atlassian Connect Express to build our Cloud app, which consists of web-facing Node JS processes and workers processes communicating through a message queue. Both types of processes use ACE for things like calling Jira’s REST API, which behind the scenes accesses the database, which in our case is PostgreSQL, to manage client keys and impersonation tokens.

Until we switched to ACE 3, we were using JugglingDB, and everything was working fine. However, once we switched to ACE 3 and thus Sequelize (since JugglingDB was removed), we started randomly getting bursts of thousands of “TimeoutError: ResourceRequest timed out” errors on our worker processes, which result in a complete loss of connectivity to our database (and an error returned by addon.settings.get) and forces us to restart the worker process. It looks a lot like the error described here:

Has anyone else experienced this?

We would gladly try the fix suggested on the aforementioned link, which consists of a different sequelize pool configuration, but unfortunately ACE doesn’t allow options to be passed to Sequelize.

Any idea?

1 Like