3 ms·
The distributed state associated with workflow systems like this makes observability and code upgrades a challenge compared to just having an application orches
by jasonthorsness 10mo ago
The distributed state associated with workflow systems like this makes observability and code upgrades a challenge compared to just having an application orchestrating workflows with progress persisted in a shared database. But if they solve those problems this would be really nice, as I've seen a ton of home-grown systems re-write checkpointing, retries, etc. and usually get it wrong.
- kondro 10mo agoThis basically is just a an application with steps that are checkpointed when they progress in a shared database (that's abstracted away from you). It's considerably simpler, less magical and cheaper than the equivalent Step Function-style implementation would be.
- jasonthorsness 10mo agoSounds like a step in the right direction. I would like to see an all-up dashboard of everything in the shared state, and good control over upgrades (maybe a mode where in-progress functions can complete on version 1, even if new functions are getting kicked off on version 2, etc.)