Lalith Suresh

The scheduler as of now is global across all worker nodes. You could hack the code to define worker nodes per federation, a scheduler for each federation, and a global...

@roymanish The DDLog Java API is used in conjunction with a custom ddlog program (that has to be loaded dynamically by your program). That already requires some out-of-band mechanisms to...

Thanks @mbudiu-vmw. I'll give this a spin.

I think we should keep it around as a draft PR for now. We'll get more data when I try out #817 in DCM (larger tables, each with more columns...

@danielhelfand @joe-kimmel-vmw just saw this. We'd be happy to look into testing kapp-controller with Sieve in the coming weeks. (cc @embano1 @marshtompsxd)

I'm leaning towards the core idea behind option 1 (the node leaving), but in a different way: I'd much rather the failure detector implementation have its own custom logic for...

Sounds good. One way to do this is to make the "should I leave logic" get a notification when an alert modifies the cut-detector (the notification can be parameterized by...

@Lbqds good catch! You are correct that the retry forever part isn't implemented yet. This piece of code here is where a process would need to keep retrying with new...

Hi @manuelbernhardt: I'd like to merge the API changes in rapid that you need to build such a broadcaster (for example, the changes to `IBroadcaster` and in the surrounding code)....

@manuelbernhardt yes, a PR with only those changes would be great. Thanks!