Back to threads
Loading user profile...
The user is running a self-hosted Supabase Edge Runtime and notices that each request seems to be handled by a fresh worker, which causes overhead for heavier functions. They are seeking guidance on worker reuse or lifecycle control to optimize performance for functions involving initialization, schema loading, and external integrations.
The user is seeking advice on managing worker lifecycle in Supabase Edge Runtime compared to self-hosted setups. They face challenges with worker initialization overhead for heavier functions and are experimenting with worker reuse. They ask for best practices and whether explicit worker reuse is advisable.
The user is seeking guidance on managing worker lifecycle and initialization in Supabase Edge Runtime compared to a self-hosted setup. They are experiencing overhead due to worker-per-request behavior and are experimenting with manual worker reuse. They seek community input on best practices and whether explicit worker reuse is considered an anti-pattern.
I'm looking for a similar alternative. If you are using Supabase Storage receiving the Supabase API directly, simply replace it with the AWS SDK, which practically solves the problem, as the calls remain almost identical. You just need to be a little careful with the region selection to obtain better performance and lower latency. Ideally, the bucket should be in the same region where the server is running. This solution using the SDK also works for other services compatible with the AWS API, such as DigitalOcean Spaces, requiring only some configuration settings.