Swarvanu Sengupta

Even if we keep a snapshot of the dag definition per request and use the snapshot to execute the flow Say a dag is D with node N1, N2, N3...

Post completion https://github.com/s8sg/goflow/issues/4, We should be able to do that without any concern, because workload will be completely independent of the DAG definition Currently it's in progress. I'll suggest to...

Still in progress. Didn’t got much time this weekend. Will try to finish it off my the next. Will require some help with the testing

@giorgiozoppi @jippi This has been requested by many. I think we can have the vanilla version using reds, with some customization option to support the queue. For a quick start...

@giorgiozoppi From project POV I'll put this in low priority. I'll focus on fix the existing issue and the bugs first. Once we have adaptability we can work on additional...

@giorgiozoppi Why we need a in memory queue, redis is in memory. Are we saying we don't need distributed approach ?

Lets pick up from the project board https://github.com/users/s8sg/projects/1 On Sat, 11 Mar 2023 at 22:57, Giorgio Zoppi ***@***.***> wrote: > We don't need a memory queue. My answer is was...

Actually flow already have Tracing where you can see elapsed time. Can you list down all the requirements and we can check in details which are the one supported and...

@giorgiozoppi This is where I need help. I didn't had enough bandwidth to look at the project recently. What I'm thinking is s simplify a lot of implementation and add...

We need to test this first https://github.com/s8sg/goflow/tree/master/core/sdk This has all core functionalities