Kyle Kingsbury

Oof, here's an even worse run. Here we appear to have gotten into some kind of full split-brain where values on different replicas diverged at some point, and the cluster...

I've got a good dozen of these cases now. Also seeing a couple new behaviors! In [this case](http://jepsen.io.s3.amazonaws.com/analyses/etcd-3.5.3/20220623T050644.000-0400.zip), clients started returning errors like `DATA_LOSS: etcdserver: corrupt cluster`, and servers also...

Hi @ahrtr! We've been tracking down bugs in lazyfs itself--still don't have all of them sanded off yet, so it's possible this may not be etcd's fault. I'm afraid my...

Hey there! Sorry I don't have this available off the top of my head, and I'm completely swamped right now with other work stuff--no longer working on etcd testing. I...

I'm afraid I can't talk too much about the present work, but I can say that the Jepsen etcd test suite has been updated to the latest jetcd and etcd,...

> not well maintained :-/

I've been maintaining Jepsen for use with physical nodes and LXC containers since 2013. I do apologize for the Docker situation--it's one of those things that people keep contributing to...

Ah, yeah, this is somewhere we could do more rigorous checking. As far as Jepsen and Elle are concerned your histories were, in fact, correct: append-only histories are *always* legal,...

Only option 1 is valid there. You want precise equality *except* for read values, which should be null in the invoke. :-)

There's no CLI argument for one-way message omissions, but I don't think it would be hard to add. Take a look at maelstrom.net for the network definition and how partitions...