Long-running Convex jobs (16MB limit + 100 item batch constraint
I'm trying to build a background job in convex where we:
- loop through a table (~20k projects)
- for each, run logic and maybe call an external API (e.g. Fly.io)
i found that convex’s migration tooling forces batches of 100 by default to avoid transaction limits
any suggestions on how to structure this whole process (db reads, marking, external calls) given that constraint and the 16mb limit?
- loop through a table (~20k projects)
- for each, run logic and maybe call an external API (e.g. Fly.io)
i found that convex’s migration tooling forces batches of 100 by default to avoid transaction limits
any suggestions on how to structure this whole process (db reads, marking, external calls) given that constraint and the 16mb limit?
