Hi Arjun, Thanks for your query. Flink is fault tolerant and supports exactly-once semantics. In your case, the aggregated values can be recovered in case of a failure or application restart. You just need to enable checkpointing and configure an appropriate state backend.
Regards, Jeyhun > > On Mon, Dec 4, 2023 at 8:03 PM arjun s <arjunjoice...@gmail.com> wrote: > >> Hello team, >> I'm relatively new to Flink's window functions, and I've configured a >> tumbling window with a 10-minute duration. I'm wondering about the scenario >> where the Flink job is restarted or the Flink application goes down. Is >> there a mechanism to persist the aggregated values, allowing the process to >> resume from where it left off after restarting? Alternatively, could you >> share some best practices for effectively handling these situations? >> >> PS : Here flink source is the File >> >> Thanks and regards, >> Arjun >> >
