Jerome Gravel-Niquet

I've rolled this out and it's nice to be able to get rid of both custom announce interval and gossip interval and have that baked in. Looks like the number...

It's looking at lot better with a a 60s `remove_down_after`! I'm wondering if the issue is that there's just too much to gossip? There are a lot of members, so...

With a 2 minutes `remove_down_after`: Can't see much because they're all overlapping 😄, now there's only a difference of 1 member for some nodes.

We often get in a situation where a single member knows about ~1-2 more/less members than the rest of the cluster. I do not know how this happens! It stays...

> If that's the case, there should be a sign of this happening: a (rare) `Error::DataFromOurselves` on your logs I can't find any such logs in the past few days...

Thanks, trying it out now, will report on how it's doing soon! As for the corrupt messages: I'm wrapping messages with my own types too. Basically sending various types of...

I think things are better since the latest release. I still see some divergence in the number of members in my cluster over time. For example, every node agreed that...

I figured as much. I think I understand you have a pretty different use case compared to ours. We do intend to use CRDTs at some point. Still designing the...

Thank you very much, that helps! I eventually figured it out, but looking at your example, maybe I should improve on the data structure a little bit. I didn't want...

[Fly.io](https://fly.io) - App hosting platform running firecracker VMs at the edge. We use Vector in production to: - Transform and send our journald logs to our elasticsearch cluster - Capture...